IRCloggy #git 2014-07-30

Logs Search ←Prev date Next date→ Channels Documentation

Provider of IRC logs since 2005.
WARNING: As Freenode became unjoinable and lost all warnings in topics, we cannot log channels on Freenode anymore.

2014-07-30

← leeN left00:00
← koi left00:00
→ Dreamer3 joined00:00
← Dreamer3 left00:00
→ sohail joined00:00
← devbug left00:01
→ Dreamer3 joined00:01
← n-st left00:01
← unomystEz left00:01
→ Tuna-Fish joined00:03
← myndzi\ left00:04
← jfelchner left00:04
← Tuna-Fish left00:04
← Not_Whatever left00:05
→ Tuna-Fish joined00:05
← jovy88 left00:05
← jay___ left00:06
← aantix left00:06
→ jovy88 joined00:06
← D-Boy left00:07
← pbrobst left00:07
jrnieder Pacsik: 'git log -g HEAD@{now}'00:08
→ veebers joined00:09
→ gargola_ joined00:09
← kristofers left00:09
← Tuna-Fish left00:09
← plasticle left00:10
← Lenton left00:10
→ micechal joined00:11
→ zumba_addict joined00:11
← _antino_ left00:11
← punchagan left00:12
→ jfelchner joined00:12
→ _antino joined00:12
→ CoderBugM joined00:13
← micechal_ left00:13
← guyz left00:14
← hahuang65 left00:14
→ maestrojed joined00:14
← mueslo left00:15
→ correcto joined00:15
← coldkey21 left00:16
← paul424 left00:16
Pacsik jrnieder: thanks much, that got me closer00:17
← CoderBugM left00:17
← smw left00:19
← lucha left00:20
→ Tuna-Fish joined00:20
→ TimmyTones joined00:20
→ kukks joined00:21
← kukks left00:21
→ kukks joined00:21
→ lucha joined00:21
← Lyqyd left00:22
← snuffeluffegus left00:22
← starkhallo left00:22
→ D-Boy joined00:23
← dreiss left00:26
→ brownbathrobe joined00:28
← brownbathrobe left00:29
zacts how can I do co-authors in a commit?00:29
← Tuna-Fish left00:29
zacts properly that is00:29
→ wakd joined00:29
zacts person a, person b, and person c developed the code for x commit.00:29
← _ikke_ left00:30
→ [ESphynx] joined00:30
→ digifiv5e joined00:30
→ _ikke_ joined00:30
digifiv5e → Guest5905500:30
← maestrojed left00:32
← ESphynx left00:32
jrnieder zacts: I want one name, email address pair to blame and contact if it goes wrong00:32
zacts: that can be one of the co-authors, or it can be a mailing list00:32
zacts ah ok, I see00:32
I think that answers my question actually..00:33
← Relsak left00:35
← Xethron left00:36
→ astrofog joined00:37
← jfelchner left00:38
← nitrix- left00:38
→ nitrix- joined00:38
nitrix- → nitrix00:38
← Guest59055 left00:38
→ Guest59055 joined00:38
→ ix007 joined00:38
← ix007 left00:38
Guest59055 → guyz00:38
← kachi8 left00:39
→ codeFiend joined00:39
→ ix007 joined00:39
→ Consuela joined00:39
← ascarter left00:39
→ Tuna-Fish joined00:40
→ jfelchner joined00:40
← correcto left00:40
← EminenceHC left00:41
→ EminenceHC joined00:41
← Tuna-Fish left00:42
← sohail left00:43
← bluesm left00:44
→ luc_ joined00:45
← luc_ left00:45
← EminenceHC left00:46
→ akushner_ joined00:48
← harish_ left00:49
→ tummy joined00:50
→ rgr joined00:51
← akushner left00:52
← akushner_ left00:53
→ Tuna-Fish joined00:53
← Tuna-Fish left00:55
← JSharpe left00:55
← _antino left00:55
← Reskp left00:55
→ Reskp joined00:56
← Willow left00:57
→ Willow joined00:58
← TimmyTones left00:59
→ TimmyTones joined00:59
← mmitchel_ left00:59
→ Nizumzen joined01:00
← Reskp left01:01
← kukks left01:02
[ESphynx] → ESphynx01:03
→ ezyang joined01:03
ezyang Does the Git mailing list actually accept unsubscribed postings?01:03
← TimmyTones left01:04
→ Destos joined01:04
→ kukks joined01:05
→ weiyang joined01:06
→ Tuna-Fish joined01:06
→ Goplat joined01:06
← zuzytel left01:07
← Vile left01:07
→ Vile joined01:08
← SethM left01:08
← m0ltar left01:08
→ _antino joined01:09
← Quevin left01:10
← Tuna-Fish left01:10
kumarat9pm Is there any command to revert back to previous state once we do checkout with out knowing previous status?01:12
like we do cd - in linux command prompt01:12
← rgr left01:12
← marcjs left01:13
→ marcjs joined01:14
→ shingshang joined01:15
← gfixler left01:16
→ k_sze[work] joined01:16
→ mizu_no_oto joined01:17
← davispuh left01:17
→ geomyidae_ joined01:18
→ coldkey21 joined01:18
← marcjs left01:19
→ hkon_ joined01:20
→ Tuna-Fish joined01:21
← axl_ left01:21
← coldkey21 left01:22
← Asiajey left01:23
← _antino left01:23
→ r0ok joined01:23
← Pacsik left01:23
→ therealGent joined01:23
← therealGent left01:23
→ therealGent joined01:23
→ axl_ joined01:23
→ causasui joined01:24
→ _antino joined01:24
← codeFiend left01:25
← hkon_ left01:25
→ pothibo joined01:25
← Tuna-Fish left01:25
→ hyperair joined01:25
→ Dougie187 joined01:26
→ Reskp joined01:27
← robustus left01:28
→ punchagan joined01:28
→ aartist joined01:29
← astrofog left01:29
← lasserix left01:29
→ harish_ joined01:30
→ robustus joined01:30
← Reskp left01:31
← ValicekB left01:31
→ mennea joined01:32
← gdrc left01:33
← awkwardusername left01:34
← Panicky left01:34
← Nizumzen left01:34
← thiago left01:36
→ el_fuego joined01:36
→ Tuna-Fish joined01:36
← mennea left01:36
→ gdrc joined01:36
→ Vampire0_ joined01:36
→ GT joined01:37
→ HumanSky joined01:38
→ daito joined01:38
→ ValicekB joined01:39
← Vampire0 left01:39
→ bezet joined01:40
← shinnya left01:40
← harish_ left01:41
→ alezandro joined01:41
← HumanSky left01:41
→ TimmyTones joined01:41
← TimmyTones left01:41
→ CoderBugM joined01:43
← mizu_no_oto left01:43
← darkbasic left01:43
→ jfelchne_ joined01:43
→ lasserix joined01:43
← jfelchner left01:43
jfelchne_ → jfelchner01:44
← CheckDavid left01:44
← bezet left01:44
→ darkbasic joined01:44
← lidaaa left01:45
← tummy left01:46
← darkbasic left01:46
→ tummy joined01:47
→ darkbasic joined01:47
← CoderBugM left01:50
← bdavenport left01:50
← tummy left01:51
← Chiftin left01:52
← msmto left01:53
→ tranzfuse joined01:54
→ jackh joined01:54
→ tokik joined01:56
jrnieder ezyang: yeah, no need to subscribe, and the usual convention is to always reply-to-all there01:56
→ megawac joined01:56
→ sohail joined01:57
→ Blacklite joined01:58
← epifanio left01:58
→ luc_ joined01:58
← FatDarrel left02:01
← ehmatthes left02:01
← thibaultcha left02:02
← Vampire0_ left02:03
→ Vampire0 joined02:03
← shruggar left02:09
← Apocalis_ left02:09
→ harish_ joined02:09
→ SLY1986 joined02:12
← Mitchbones left02:13
← Noldorin left02:15
→ nkouevda joined02:16
← binaryhat_ left02:16
→ binaryhat joined02:17
← TDJACR left02:18
← axl_ left02:18
← kevinb left02:19
→ TDJACR joined02:19
→ kevinb joined02:19
← Arkaniad left02:20
→ tummy joined02:20
→ deuti joined02:21
→ axl_ joined02:21
← darkxploit left02:22
→ otter768 joined02:22
→ diegoviola joined02:24
→ codeFiend joined02:24
← MickeyO left02:25
← lymmz left02:27
→ Reskp joined02:27
← aartist left02:29
← codeFiend left02:30
← luc_ left02:30
← Tuna-Fish left02:30
→ luc_ joined02:31
← LucasTT_ left02:31
→ Tuna-Fish joined02:31
→ LucasTT joined02:31
← LucasTT left02:32
← Reskp left02:32
→ mennea joined02:32
← Hello71 left02:33
← Tuna-Fish left02:36
→ ropav joined02:37
← mennea left02:37
→ curiousmansi_ joined02:40
← tummy left02:40
← tranzfuse left02:40
← synchronous left02:40
← Consuela left02:41
→ tummy joined02:41
→ hahuang65 joined02:41
→ codeFiend joined02:41
→ synchronous joined02:42
→ EminenceHC joined02:42
→ Arkaniad joined02:43
→ Embassy joined02:45
← tummy left02:45
← EminenceHC left02:47
→ Tuna-Fish joined02:47
← a_le left02:47
← Embassy left02:50
→ tatsuya_o joined02:51
→ dogeydogey joined02:52
← tadni` left02:52
← bitkiller left02:52
← harish_ left02:55
→ yacks joined02:55
← tatsuya_o left02:55
← Tuna-Fish left02:56
→ Tuna-Fish joined02:57
← Destos left02:57
← ISmithers left02:57
→ fatalhalt joined03:00
→ Reskp joined03:03
← tokik left03:05
→ knighthawk joined03:05
← Reskp left03:07
← kermit left03:07
→ thibaultcha joined03:08
→ kermit joined03:08
→ ascarter joined03:08
→ bdavenport joined03:08
→ arup_r joined03:09
← arup_r left03:11
→ harish_ joined03:11
→ arup_r joined03:11
← kerrick left03:11
← eliasr left03:12
← akim__ left03:12
← deanclkclk left03:13
← moshR left03:14
→ kerrick joined03:14
← arup_r left03:15
← kerrick left03:15
→ kerrick joined03:17
← kerrick left03:18
← Peuc left03:18
← solker left03:19
→ hkon_ joined03:21
← ORiON-- left03:21
→ jimmy_lo joined03:22
← knownasilya left03:23
← netj left03:23
→ moshR joined03:23
→ netj joined03:23
→ ORiON-- joined03:23
← _antino left03:24
← hkon_ left03:26
← tplaner left03:26
→ thiago joined03:28
→ tplaner joined03:28
← hahuang65 left03:29
← jimmy_lo left03:29
→ tokik joined03:30
← ORiON-- left03:31
→ ORiON-- joined03:31
→ natechan joined03:32
→ jimmy_lo joined03:32
← natechan left03:32
← jimmy_lo left03:33
→ sohail_ joined03:34
← sohail left03:36
← axl_ left03:36
← pandeiro left03:37
→ epifanio joined03:38
← ORiON-- left03:40
→ bezet joined03:40
→ ORiON-- joined03:42
← alvaro_o left03:44
← bezet left03:44
← tokik left03:45
← otter768 left03:46
← Dougie187 left03:46
← tplaner left03:49
→ axl_ joined03:50
→ quazimodo joined03:50
← alezandro left03:51
→ tatsuya_o joined03:51
→ FatDarrel joined03:51
← sohail_ left03:51
← tatsuya_o left03:55
→ pweaver_ joined03:55
← FatDarrel left03:56
→ FatDarrel joined03:56
→ Consuela joined03:58
← ORiON-- left03:58
← Consuela left03:59
→ ORiON-- joined04:00
→ pnsrc joined04:01
→ tummy joined04:03
→ JohnBat26 joined04:04
← m0viefreak left04:04
→ tatsuya_o joined04:04
→ alezandro joined04:06
← tatsuya_o left04:09
← diegoviola left04:09
← lazers left04:10
← quazimodo left04:11
→ diegoviola joined04:11
→ quazimodo joined04:11
← cwillu_at_work left04:13
→ curiousmansi__ joined04:14
← veebers left04:14
← curiousmansi_ left04:15
← therealGent left04:16
→ mennea joined04:16
→ NBhosting joined04:16
← macabre__ left04:17
→ macabre joined04:18
← gargola_ left04:18
→ gargola__ joined04:18
← deuti left04:19
→ a_le joined04:20
← tummy left04:21
← kukks left04:21
→ tummy joined04:22
→ foist joined04:22
← hellome left04:22
foist Is there some way for me to execute my tests on a commit that isn't currently checked out?04:22
milki what are your tests?04:23
foist Some python tests.04:23
milki and what is required for those tests?04:23
→ Mobutils joined04:23
← dogeydogey left04:23
foist The environment doesn't need to change - just the codebase.04:24
→ Tommy[D] joined04:24
milki i dont see how that answers my question. what do you mean?04:24
← daito left04:24
foist I'm not sure I understand your question. What is required? It's a simple test along the lines of `assert var == 2`04:25
milki uh04:25
so you have a test file04:25
foist Yes.04:26
← ropav left04:26
milki what do you need to make it run successfully?04:26
you mention environment and codebase04:26
are those two things required for your test?04:26
← tummy left04:26
foist I just execute the test command `./manage.py test`04:26
milki ok04:26
foist Yes, those things are.04:26
← Tommy[D]_ left04:26
milki so, so if you do that, does it run successfully without any thing else?04:27
→ cwillu_at_work joined04:27
foist Yes.04:27
milki then i dont see why you need a checkout04:27
← toedwy left04:27
foist Lets say I'm on commit B. It's different from commit A. I want to compare test results from both commits. Is it possible to do so without moving my HEAD?04:28
milki ah04:28
sort of04:29
i see a couple options04:29
← luc_ left04:29
milki hmm04:29
→ luc_ joined04:29
milki you can archive the commit to a different location04:29
and then copy your tests over there to run them04:29
or you can checkout the commit and then checkout the tests from HEAD04:30
or you can have HEAD as working dir, and checkout the codebase from the commit04:30
→ whyy joined04:30
milki but your tests need to run with actual files from commit B04:31
so you need to somehow get those files on disk before you can run your tests04:31
foist I see - it's more trouble than it's worth. I thought there may be some trickery that could make it quick-n-dirty.04:32
milki perhaps04:32
← whyy left04:33
→ Kicer86 joined04:33
→ whyy joined04:34
→ beautifulmind joined04:35
foist Alright. Thanks, milki.04:35
← megawac left04:36
← robbyoconnor left04:37
→ r0bby_ joined04:37
← r0bby_ left04:39
→ veebers joined04:41
→ lidaaa joined04:41
→ pikaren joined04:41
→ EminenceHC joined04:43
← pnsrc left04:43
← macabre left04:43
← beautifulmind left04:43
→ beautifulmind joined04:44
→ dreiss joined04:44
← dreiss left04:44
← the_cybersphinx left04:45
→ dreiss joined04:45
← epifanio left04:45
→ gfixler joined04:46
→ Embassy joined04:47
← EminenceHC left04:47
← DeathIncarnate left04:47
→ Textmode joined04:47
→ mdel joined04:48
← fatalhalt left04:49
→ hj2014 joined04:51
← Embassy left04:51
← ascarter left04:52
← northfurr left04:52
← Gamecubic left04:52
← JohnBat26 left04:53
→ northfurr joined04:53
→ hahuang65 joined04:56
← fifo left04:57
→ maveneagle joined04:57
← HanClinto left04:58
← dreiss left04:59
← joelm left04:59
→ UmbraMalison_ joined05:02
← UmbraMalison left05:03
← unglid left05:04
← curiousmansi__ left05:05
→ plitzenberger joined05:06
← mkoskar left05:07
→ JakeSays joined05:08
→ zz_unglid joined05:09
← zumba_addict left05:10
JakeSays ok so i have a really weird problem that i cant seem to work through. earlier in the day i was merging from origin. there were conflicts and a dialog appeared (i'm using git extensions UI). at that very moment my wife interrupted me and my dog jumped on the keyboard. dialog went away and now my repo is f*'d up. at the top of my commit history i have 'commit index' and i have no idea what to do with it :(05:10
→ zumba_addict joined05:10
← plitzenberger left05:11
← diegoviola left05:11
← maveneagle left05:11
→ diegoviola joined05:12
← punchagan left05:12
→ stodan joined05:12
→ johnkevinmbasco joined05:12
← Mobutils left05:12
milki lololol05:12
uh05:12
→ maveneagle joined05:12
milki are you still in a merge?05:13
← veebers left05:13
→ Mobutils joined05:13
← axl_ left05:13
JakeSays milki: i dont think so05:13
→ axl_ joined05:14
JakeSays milki: how would i tell?05:14
← hj2014 left05:15
→ pantsman_ joined05:16
→ hj2014 joined05:16
← quazimodo left05:18
← johnkevinmbasco left05:18
← zivester left05:19
→ johnkevinmbasco joined05:19
milki git status05:19
or the presence of a MERGE_HEAD file in .git05:19
JakeSays no merge_head05:20
and git status shows me the list of staged files to be committed05:20
but i dont know how to get back to the point where i can resolve the conflicts05:21
milki JakeSays: hm, then look at git reflog05:21
you can git reset --hard to the commit before the merge05:22
and then try the merge again05:22
→ daito joined05:22
→ hkon_ joined05:22
JakeSays but then i lose my changes05:22
→ newbie999 joined05:23
milki well, if you dont know what state you are in, thats not such a bad outcome05:23
JakeSays oh it is - its an entire day's work05:23
milki if you want to preserve something but still be in a merge, i dont know05:23
JakeSays hmm. wonder if i can stash05:24
→ kaendfinger joined05:25
kaendfinger I am writing a GitHub Client Library for Dart, and I figured I would share this here: http://bot.directcode.org:8080/browser/repository.html?user=git&sort=stars05:26
← hkon_ left05:26
ovidiu-florin canton7: are you still here? sorry, I haven't seen your responce yesterday. Yes, I readdly do branch TRUNK/my/subdir/somefolder to BRANCHES/mybranchsDir/branch105:26
→ joelm joined05:27
ovidiu-florin s/readdly/really05:28
→ kyb3r_ joined05:28
→ knownasilya joined05:28
→ certainstrings joined05:29
→ johnkevi_ joined05:29
← beautifulmind left05:30
→ awkwardusername joined05:30
← johnkevinmbasco left05:30
→ Balliad joined05:31
→ arup_r joined05:31
← joelm left05:32
→ tharvey_ joined05:33
← arup_r left05:33
← Megaf left05:34
→ johnkevinmbasco joined05:34
→ arup_r joined05:34
← johnkevi_ left05:35
← Balliad left05:35
← pantsman_ left05:35
tharvey_ what options would I have for setting up multiple git+ssh accounts on a single server? What I'm after is that I need several secure repositories but would like to not setup mulitple servers - can I do this on a repo path basis or do I need multiple hosts or ports?05:36
→ aphirst joined05:36
certainstrings evening all05:36
→ lite_ joined05:36
milki tharvey_: !gitolite05:36
gitinfo tharvey_: 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/05:36
aphirst Ren & Stimpy reference? I'm loving this channel already.05:36
Does anyone know off the top of their head which syntax highlighting library is used by `gitg`'s code preview?05:37
← awkwardusername left05:37
← Samual left05:37
→ arup_r_ joined05:37
← axl_ left05:37
aphirst I want to update its list of Fortran keywords/functions05:37
milki havent ever heard of gitg05:38
← arup_r left05:38
→ toedwy joined05:38
tharvey_ milki: thanks that looks perfect!05:38
milki sitaram: ^ you have another client!05:38
aphirst Hang on, isn't `gitg` included in GNU/Linux installed of git? Or have I made a mistake?05:39
osse aphirst: wild guess: https://wiki.gnome.org/Projects/GtkSourceView - it's what gedit uses05:39
→ kcj joined05:39
aphirst osse: I think you might be right, thanks a bunch, I'll go poke them05:40
← kcj left05:40
aphirst Cheers, guise.05:40
← aphirst left05:41
→ SethM joined05:41
→ St_Marx joined05:42
← salamanderrake left05:42
→ FilipNortic joined05:43
← toedwy left05:43
← GT left05:43
← certainstrings left05:44
← maveneagle left05:45
→ curiousmansi_ joined05:45
→ deanclkclk joined05:46
→ maveneagle joined05:46
← maveneagle left05:46
→ ThomasLocke joined05:46
→ maveneagle joined05:47
← maveneagle left05:47
← foist left05:47
→ Samual joined05:48
← Samual left05:48
→ Samual joined05:48
→ maveneagle joined05:48
← el_fuego left05:49
← visceral left05:50
← Tuna-Fish left05:51
← curiousmansi_ left05:51
→ smithbone joined05:52
← smithbone_ left05:53
← arup_r_ left05:53
← madewokherd left05:53
→ arup_r joined05:54
← arup_r left05:54
→ arup_r joined05:54
sitaram milki: :)05:55
→ chaos7theory joined05:56
← ValicekB left05:56
← Brando753 left05:57
→ LukeT- joined05:57
→ Brando753 joined05:58
← thibaultcha left05:58
→ axl_ joined05:58
→ beautifulmind joined05:59
← deanclkclk left05:59
← beautifulmind left06:00
→ quazimodo joined06:00
← axl_ left06:01
→ koi joined06:01
← LukeT- left06:02
→ LukeT- joined06:02
→ axl_ joined06:02
→ Tuna-Fish joined06:02
LukeT- → LukeT06:02
→ Wiz_KeeD joined06:03
→ ValicekB joined06:04
→ MattMaker joined06:05
→ Reskp joined06:05
→ papegaaij joined06:05
→ Akiva-Thinkpad joined06:06
→ cilly joined06:06
→ mlechner joined06:07
→ Kulrak joined06:07
← Kulrak left06:07
→ Kulrak joined06:07
Akiva-Thinkpad Can git be setup to do two way sync? I am developing a pair programming plugin, and I am looking into different sync frameworks06:08
← kumarat9pm left06:08
← Reskp left06:10
→ freimatz joined06:10
→ cbreak-work joined06:10
← geomyidae_ left06:11
→ hoschi joined06:11
→ KevinTS joined06:11
milki Akiva-Thinkpad: its not recommended. its more command to have a client-server model in which you have a single bare repo and both clients talk to that repo06:14
→ Xethron joined06:16
← Xethron left06:16
→ Xethron joined06:16
thiago Akiva-Thinkpad: one person can pull from the other06:16
Akiva-Thinkpad thiago, and milki thank you for explaining that.06:17
← montchr left06:17
Akiva-Thinkpad brb06:17
← a_le left06:20
Akiva-Thinkpad milki, would you mind doing a small favour? I opened this stack exchange question here: http://askubuntu.com/questions/504413/what-is-the-fastest-two-way-sync-framework-available-on-ubuntu , and I am collecting answers from possible solutions. It would mean a lot to me if you could provide a brief synopsis of pros and cons to using Git for two way sync.06:20
← northfurr left06:20
Akiva-Thinkpad however in respect to your time, feel free to decline06:20
← dcpc007 left06:20
→ dcpc007 joined06:21
milki doesnt use SO06:22
Akiva-Thinkpad :o06:22
← FatDarrel left06:22
→ Reskp joined06:23
Akiva-Thinkpad Well then to the broader community; if anyone wishes to help me in this regard, it would be both upvoted, and greatly appreciated.06:24
gitinfo set mode: +v06:24
→ Errorical joined06:24
lidaaa Akiva-Thinkpad, I never tried it but there is this https://www.youtube.com/watch?v=QCDEgvz_Rqw06:24
* #git :Cannot send to channel06:24
gitinfo set mode: +v06:25
lidaaa Akiva-Thinkpad, I never tried it but there is this https://www.youtube.com/watch?v=QCDEgvz_Rqw06:25
→ autochthon joined06:25
Akiva-Thinkpad lidaaa, Yah I have seen that. They are using a central server they host to collaberate though.06:25
I'd like to see what framework they are using, but I have not been able to get a hold of any of them in their irc channel06:25
← tpollard left06:26
lidaaa doesn't look like you need git, you want a realtime collaboration editing solution06:26
→ rupen joined06:26
← Mobutils left06:27
→ sinkensabe joined06:27
lidaaa which will need a server to sync the clients changes06:27
Akiva-Thinkpad lidaaa, I am afraid you have me at a disadvantage06:28
lidaaa, are you suggesting something such as VNC?06:28
lidaaa, I am flowcharting my design at the moment; perhaps this will illustrate better what I am trying to accomplish: https://docs.google.com/spreadsheets/d/1Hv6GY0SjyGreuuNziK_1HmTJy6p69gmyNtjFV4yxiVk/edit#gid=53576894706:30
→ beautifulmind joined06:30
lidaaa something like vnc, will be one solution. but better then that is having a solution like gobby for example. there is also toghterjs by mozilla06:30
google docs does it nicely too06:31
Akiva-Thinkpad lidaaa, Do you understand that I am developing a plugin for a specific ide?06:31
← beautifulmind left06:31
→ hinogi joined06:31
→ zoso joined06:31
→ Junior joined06:31
→ Mobutils joined06:32
→ beautifulmind joined06:32
lidaaa doesn't change the fact that you need to build/integrate a collaboration protocol06:32
Akiva-Thinkpad The goal is to utilize the advantages of the ide, that would be lost if one simply used google docs to collaboratively edit code.06:32
← jayrobot left06:32
lidaaa I reference google docs, but I am refering more to the underlying technology.06:33
→ Macaveli joined06:33
Akiva-Thinkpad lidaaa, right...06:33
lidaaa, I apologize; you still have me at a disadvantage. I am looking for an underlying framework which to work with.06:34
→ B1x1t0 joined06:34
← lucha left06:35
Akiva-Thinkpad I do not understand how referencing google docs or gobby without mentioning what framework it is using, answers my question. I am not saying it doesnt, I am just not an expert when it comes to sync frameworks.06:35
lidaaa I started by saying that you are not looking for a git solution. and you need to find something that is more suitable for realtime editing, but I guess you know that by now06:36
let say you sync files using git, how are you going to update the ide with the new content?06:36
→ phanimahesh joined06:37
gitinfo set mode: +v06:37
Akiva-Thinkpad lidaaa, Good question.06:37
→ LeMike joined06:37
→ Embassy joined06:37
→ Gege joined06:37
Akiva-Thinkpad lidaaa, What I will have to do inevitably is change the ide's settings to do a realtime update, instead of prompting the user when a file has changed.06:37
osse what about the user's own unsaved changed?06:38
changes06:38
Akiva-Thinkpad However, that is a question for the #qt;06:38
→ a_le joined06:38
lidaaa and let say the file you are auto reladoing is big how you are going to maintaine the right scroll position06:39
← a_le left06:39
← Reskp left06:39
→ zumba_ad_ joined06:39
Akiva-Thinkpad osse, Good question. To that, the other thing I will edit in the ide, is that all changes, even a character input, are immediately saved.06:39
← Tuna-Fish left06:39
osse Gobby uses libinfinity, but it's a library you link against, not a framework.06:39
→ Reskp joined06:39
osse or maybe a framework is a library06:39
lidaaa the child has many names06:39
→ ehmatthes joined06:39
Akiva-Thinkpad osse, To be honest; I am not sure of the technical distinction between a library and a framework.06:39
lidaaa I suggest you change the name library or framework to "the solution I need"06:40
osse framework sounds very web-2.0 to me :P library sounds like too much work06:40
Akiva-Thinkpad lidaaa, You ask many good questions; however this is red herring to the topic at hand.06:40
lidaaa what I am trying to get to, is that basic file syncing is not proper for real time collaboration06:41
← jeffhung left06:41
→ donnib joined06:42
← zumba_addict left06:42
← rupen left06:42
Akiva-Thinkpad lidaaa, Well thank you for your input. I acknowledge that there are hurdles to this project. However; maintaining the scroll bar position with a synced file, is not a reason against using basic file syncing.06:42
that is a question of ide configuration.06:43
→ EminenceHC joined06:44
← cilly left06:44
← harish_ left06:45
← MattMaker left06:46
→ cxz joined06:46
← cxz left06:46
→ cxz joined06:46
hinogi Hello, I run in some difficulties trying to use git svn clone with a path to network drive or even a local folder06:46
← Reskp left06:46
→ MattMaker joined06:47
→ Reskp joined06:47
← polyzen left06:47
→ jeffhung joined06:47
hinogi Couldn't open a repository: Unable to open an ra_local session to URL: Unable to open repository 'file:///D:/Testing/tmp/test-svn': Expected FS format '2'; found format '6' at /usr/lib/perl5/site_perl/Git/SVN.pm line 14806:47
← MattMaker left06:47
hinogi but svn ---version has ra_local module and ra_serf06:48
osse all the collaborative editors I've seen don't do tricks to copy files over the network. Everything is done in memory06:48
← reaverb left06:48
→ foist joined06:48
← EminenceHC left06:48
← chaos7theory left06:48
→ jargon joined06:49
← arup_r left06:49
→ chaos7theory joined06:49
→ jnavila joined06:49
→ arup_r joined06:49
→ tatsuya_o joined06:49
→ Tuna-Fish joined06:50
→ n-st joined06:50
→ hkon_ joined06:51
→ lb1a joined06:51
→ tvw joined06:51
← Reskp left06:52
← hoschi left06:52
← Goplat left06:52
← foist left06:53
lidaaa Akiva-Thinkpad, This is by no mean a simple project, You probably can get something working using git. with alot of complexities and extra work, or you can start with something that was designed to solve your particualr need which is realtime collaboration.06:53
I did mention TogetherJS by mozilla. JSFiddle also support it, you can see the collaboration button06:54
Akiva-Thinkpad lidaaa, Right. I am looking for all possible solutions, trying to determine the pros and cons, and will make a decision from there.06:54
TogetherJS... is this for collaboratively developing javascript?06:55
lidaaa no it is for collaborative web apps06:55
→ mishok13 joined06:55
Akiva-Thinkpad lidaaa, So, again, this is for an ide.06:55
not a web application.06:55
lidaaa which ide?06:55
Akiva-Thinkpad Qt Creator.06:56
aka06:56
→ aidalgol joined06:56
Akiva-Thinkpad Ubuntu SDK.06:56
← stfn left06:56
← ehmatthes left06:56
→ Balliad joined06:56
→ gua_ould joined06:57
→ arup_r_ joined06:57
→ harish_ joined06:58
→ hebz0rl joined06:59
→ stfn joined06:59
→ bvi_ joined07:00
Akiva-Thinkpad lidaaa, Actually, there is something you could do to greatly help me and the broader community, if you are well familiar with Git. I posted this Stack Exchange Question regarding two way sync frameworks, and am collecting answers from all possible solutions, with their pros and cons. I would honestly be greatly appreciative if you or anyone else here contributed an answer for git.07:00
http://askubuntu.com/questions/504413/what-is-the-fastest-two-way-sync-framework-available-on-ubuntu07:00
→ beautifulmind1 joined07:00
← arup_r left07:00
→ salamanderrake joined07:00
← bvi left07:00
Akiva-Thinkpad Of course in respect to your time, it is perfectly fine for you to decline.07:01
→ ErKa joined07:01
← beautifulmind left07:01
→ johnkevi_ joined07:01
← mdel left07:02
← johnkevinmbasco left07:02
→ ferreum joined07:03
← codeFiend left07:03
→ lucha joined07:03
← lucha left07:03
← harish_ left07:05
lidaaa I am by no mean git expert, but from the little I know about git, I know that this is not what you are looking for. I don't mind answering your questions here. as to why not to use git07:05
← bvi_ left07:05
→ johnkevinmbasco joined07:06
← johnkevi_ left07:06
→ plitzenberger joined07:06
← beautifulmind1 left07:07
→ ghostlines joined07:10
→ johnkevi_ joined07:10
← plitzenberger left07:11
← mennea left07:11
→ beautifulmind joined07:11
← johnkevinmbasco left07:11
→ solker joined07:11
→ _tpavel joined07:12
lidaaa Git is not meant for realtime collaboration. you can also check the apache wave project based on google wave protocol http://wave-protocol.googlecode.com/hg/spec/federation/wavespec.html07:12
→ jluc joined07:12
Akiva-Thinkpad lidaaa, Thank you; I really appreciate your input.07:13
← alezandro left07:14
moritz in particular, git is meant to work with discrete snapshots of your files (called commits); that's not what you want in real-time collaboration07:14
→ sdrfnord1 joined07:15
Akiva-Thinkpad moritz, right; its using project states rather than individual file states; is that what you are basically driving at?07:15
→ jagob joined07:15
→ kriks joined07:15
→ endiruna joined07:16
← beautifulmind left07:16
← johnkevi_ left07:16
→ hamster joined07:17
← jonasdk89 left07:17
→ jonasdk89 joined07:17
moritz Akiva-Thinkpad: that, and that those commits have commit messages which you must manually enter07:17
Akiva-Thinkpad: and which imply that people use them for more coarse changesets than you'd want in real-time collab07:17
*implies07:18
← nkouevda left07:18
Akiva-Thinkpad moritz, you would not be able to setup git to forego commit messages?07:18
→ nkouevda joined07:18
moritz Akiva-Thinkpad: well, you could auto-generate them, but they'd be pretty useless07:19
you'd basically have to hijack the system07:19
and if you're hijacking a system, it's good to consider if there's a system you could use that you don't have to hijack07:19
← sinkensabe left07:19
Akiva-Thinkpad moritz, that is a great point.07:20
to my earlier link, would you care to offer an answer to Git's viability?07:20
http://askubuntu.com/questions/504413/what-is-the-fastest-two-way-sync-framework-available-on-ubuntu/504444#50444407:20
→ psoo joined07:20
→ johnkevinmbasco joined07:21
moritz Akiva-Thinkpad: git isn't meant for that07:22
← psoo left07:22
moritz Akiva-Thinkpad: I'd go with unison or something like aerofs07:22
→ psoo joined07:22
moritz unison isn't real-time either07:22
← nkouevda left07:22
Akiva-Thinkpad moritz, Yes; I am looking into unison at the moment. I have never heard of Aerofs.07:23
moritz (aerofs is basically a dropbox clone, but you have more control over your data)07:23
haven't tried it yet though07:23
Akiva-Thinkpad moritz, despite Git not being the best answer for that thread, I think it would still be valuable to leave an answer there. Git from what I understand is incredibly fast in syncing.07:23
→ sea-gull_ joined07:24
moritz Akiva-Thinkpad: there's syncinc and there's syncing07:24
← Tuna-Fish left07:24
→ BunnyHolder joined07:24
lidaaa and there is sinking07:24
Akiva-Thinkpad I do not understand why, but from what I gather, it served the linux kernel much better than cvs or tarballs.07:24
moritz but the linux kernel isn't real-time synced either07:25
Akiva-Thinkpad moritz, Well think of it this way; if you leave an answer there, then no one will come back to #git asking again :P Thats what I love about stack exchange.07:25
→ jottr joined07:25
moritz Akiva-Thinkpad: stack exchange is meant to have the "one true answer" for questions, so answering with what not to use is basically off-topic there07:26
(I don't like this model myself, but I do try to play by their rules)07:26
→ dotpot joined07:26
Akiva-Thinkpad moritz, Well the question is specifically about speed; the quickest solution.07:27
→ numdig joined07:27
Akiva-Thinkpad moritz, in the case of the linux kernel, it is "The Quickest solution".07:27
moritz Akiva-Thinkpad: no07:27
Akiva-Thinkpad so I think the answer will be valuable.07:27
moritz Akiva-Thinkpad: it's the solution that's best suited for the linux development model07:27
← johnkevinmbasco left07:27
← ErKa left07:27
→ justCarakas joined07:27
moritz Akiva-Thinkpad: and that specifically includes manual choice about which parts to sync07:28
Akiva-Thinkpad: which is *not* what the question is about07:28
← Mobutils left07:28
→ SonderbladeWork joined07:28
→ vital joined07:28
← SonderbladeWork left07:28
→ SonderbladeWork joined07:29
→ vurma joined07:29
← cxz left07:30
→ freezevee joined07:31
← gargola__ left07:31
Akiva-Thinkpad moritz, You are correct; I do not want to take that away from you, however it is still valuable to answer. I am currently discussing in Meta whether answers that advocate against possible solutions should be included.07:31
moritz, also, http://meta.askubuntu.com/questions/3803/should-i-answer-a-question-which-already-has-an-accepted-answer?rq=107:31
← Xethron left07:31
lidaaa collaboration is more complex then I had in mind http://en.wikipedia.org/wiki/Operational_transformation07:31
Akiva-Thinkpad According to this meta se discussion, "If you have something new to offer that answers the question and is not appropriate as an edit to another answer, then you are helping the community (and sometimes even the OP) by posting it as a new answer."07:31
← jottr left07:32
Akiva-Thinkpad lidaaa, yah; no kidding :P07:33
→ Tuna-Fish joined07:35
→ esclear joined07:37
lidaaa oh well this is an intersting topic, but it is late here I can't think anymore. good night07:38
→ hoschi joined07:38
Akiva-Thinkpad lidaaa, good night. Thanks again.07:38
→ ErKa joined07:40
lidaaa np , btw Akiva-Thinkpad, I would also suggest you add to the question you posted to reflect the realtime / collaborative nature.07:41
→ bezet joined07:41
Akiva-Thinkpad lidaaa, I am considering it. I am undecided, but I think that is a reasonable suggestion.07:42
← Vile left07:42
← pweaver_ left07:43
← jagob left07:44
→ Sceorem joined07:44
freezevee I fucking love git people. I just wanted you all to know. I love it. I want to marry it.07:44
→ sinkensabe joined07:44
freezevee I needed to say this to people that will understand me.07:44
lng Hi! When I run `git checkout master; git fetch origin master; git rebase origin/master`, I get "Current branch master is up to date.", but `git pull --rebase` is updated master: "ac5227f..8f3d5c3 master -> origin/master". What is the difference? Sometimes the first approach works, some other time, doesn't...07:44
→ nour joined07:44
← nnyk_ left07:45
grawity lng: the second one is like "git fetch origin && git rebase origin/master"07:45
lng grawity: yes07:45
← bezet left07:45
→ ColdKeyboard joined07:45
grawity lng: the difference being that it always downloads the latest stuff from origin07:45
← justCarakas left07:45
grawity meanwhile, "origin/master" *always* means the local, cached, copy of the 'master' branch that was fetched from origin last time.07:46
lng grawity: but why doesn't it download latest commits in the first command set?07:46
git fetch origin master should do it07:46
grawity hmm07:46
ah, right07:46
lng yea07:46
and it is not always like that07:46
grawity in the 4-word version, `git fetch` does not fetch all branches, and does not update the remote-tracking branches (the cached copies)07:46
→ pweaver_ joined07:47
grawity instead, it fetches the given branch into a special ref "FETCH_HEAD"07:47
gfixler nods knowingly at freezevee07:47
lng git fetch origin master;07:47
grawity this has been fixed in Git 2.0 or one of the recent 1.x's, but it used to trick many people before07:47
so much that !fetch4 is a factoid here07:48
gitinfo [!fetchfour] [pre 1.8.4 only] We recommend against using 'git fetch/pull <remote> <refspec>' (i.e. with branch argument), because it doesn't update the <remote>/<branch> ref. The easy way to fetch things properly is to get everything: 'git fetch' or 'git pull' are sufficient if you have one remote; otherwise we recommend 'git fetch <remote>' (plus 'git merge <remote>/<branch>' if you wanted to pull/merge).07:48
grawity in other words, do *not* `git fetch origin master`; do `git fetch origin` instead07:48
lng interesting07:49
→ thevishy joined07:50
→ dsdeiz_ joined07:50
grawity yeah, fixed in 1.8.4, but it's still not worth using it07:50
lng I'm on 1.7.9.507:51
that's why07:51
so07:51
→ Mobutils joined07:52
lng `git checkout master; git fetch origin; git rebase origin/master`?07:52
← dsdeiz left07:52
grawity yeah07:52
lng is the last part okay?07:52
grawity yes07:52
lng or just git pull --rebase which is shorter07:53
grawity yeah – `git checkout master; git pull --rebase`07:54
lng yea07:54
thanks07:54
was it a bug prior 1.8?07:54
grawity not really07:54
lng a feature? :-)07:54
← arup_r_ left07:55
→ W0rmDr1nk joined07:55
grawity kinda – I think it was *meant to* fetch a given branch into FETCH_HEAD without modifying anything else07:55
perhaps for scripting07:55
→ arup_r joined07:55
→ z8z joined07:55
← Tuna-Fish left07:57
→ intellix joined07:58
← lidaaa left07:58
→ mikecmpbll joined07:59
→ zequence_ joined07:59
← zequence left07:59
→ Learath2 joined07:59
← arup_r left07:59
gitinfo set mode: +v07:59
→ marduk191 joined08:00
→ zapotek6 joined08:00
← zapotek6 left08:00
← chaos7theory left08:00
marduk191 is there a way to clone a branch from one remote repo to a branch in my local repo and still keep the history?08:02
← pikaren left08:02
→ zumba_addict joined08:02
osse marduk191: heh, it takes more effort to *not* keep the history :)08:02
→ jackneill joined08:03
→ zapotek6 joined08:03
osse marduk191: "clone a branch" doesn't exactly fit together though. Do you have a clone already and what to retrieve a branch from where you cloned from, or from some other repo ?08:03
← zumba_ad_ left08:03
marduk191 alright we are working on kernel branches, my past 3 branches are in a completely dirrerent repo. I just want to pull in their branch and then sync it as a new branch in my repo for now08:04
→ Tuna-Fish joined08:04
marduk191 push08:05
← whyy left08:05
marduk191 lol I'm a noob to git08:05
→ zumba_ad_ joined08:05
→ pikaren joined08:06
← pweaver_ left08:06
← zumba_addict left08:08
→ shruggar joined08:08
→ johnkevinmbasco joined08:08
osse marduk191: "their"?08:08
marduk191 the other repository08:09
→ marcjs joined08:10
osse marduk191: have you added it as a remote?08:10
marduk191 I haven't done anything yet, I thought I'd ask. My idea was to git remote add name <url>, then git fetch name, then git checkout -b name branch name/branch08:12
osse marduk191: that sounds 100% correct08:12
marduk191 that won't give me staging status or anything?08:12
→ mennea joined08:13
← knownasilya left08:13
marduk191 in master08:13
eh I'm gong for it. I'll cross that when I get there lol. Thanks08:14
→ mueslo joined08:14
← ezyang left08:14
→ ezyang joined08:14
→ plitzenberger joined08:14
ezyang jrnieder: So, if I make a post and it doesn't show up in the archives, who do I bug?08:14
→ calculon joined08:15
← zapotek6 left08:16
→ GoProDeki joined08:16
→ martinklepsch joined08:16
→ zapotek6 joined08:16
← GoProDeki left08:16
← zapotek6 left08:17
→ Novitsh_ joined08:17
→ r0bby_ joined08:17
← shruggar left08:17
→ rob_w joined08:18
← daito left08:19
→ zapotek6 joined08:19
← zapotek6 left08:19
→ msmto joined08:20
→ tokik joined08:20
→ Droolio joined08:20
→ correcto joined08:21
r0bby_ → robbyoconnor08:23
→ whyy joined08:23
→ arup_r joined08:23
← johnkevinmbasco left08:23
→ zapotek6 joined08:24
← zapotek6 left08:24
→ qt-x joined08:24
← martinklepsch left08:25
← thevishy left08:25
← pikaren left08:25
→ Sound joined08:25
← marcjs left08:26
→ marcjs joined08:26
→ shane1 joined08:26
← weiyang left08:27
← watabou left08:27
← endiruna left08:27
← Mobutils left08:27
← Rhonda left08:29
→ atamanroman joined08:30
← Willow left08:30
→ Willow joined08:30
← intellix left08:30
→ Mobutils joined08:30
← marcjs left08:30
← axl_ left08:31
→ weiyang joined08:31
← thiago left08:32
← plitzenberger left08:32
→ shruggar joined08:33
→ Rhonda joined08:33
← bgerber left08:34
← Learath2 left08:34
→ intellix joined08:34
→ Learath2 joined08:34
← Willow left08:34
gitinfo set mode: +v08:35
← Rhonda left08:35
→ Rhonda joined08:35
→ Xethron joined08:36
← Xethron left08:36
→ Xethron joined08:36
→ bezet joined08:36
→ jayrobot joined08:37
← Learath2 left08:37
→ dawkirst joined08:37
→ Learath2 joined08:37
gitinfo set mode: +v08:38
← n-st left08:38
→ rgr joined08:38
← oojacoboo left08:38
→ oojacoboo joined08:39
→ die joined08:40
→ diegovio1 joined08:41
← die left08:41
→ die joined08:42
← Mobutils left08:42
← die left08:42
← correcto left08:42
diegoviola → Guest4342508:42
diegovio1 → diegoviola08:42
← Guest43425 left08:43
← SLY1986 left08:43
→ EminenceHC joined08:44
← whyy left08:46
← Akiva-Thinkpad left08:47
← EminenceHC left08:49
→ nnyk_ joined08:49
← rgr left08:49
marduk191 alright, that was perfect08:50
→ daito joined08:51
← calculon left08:51
→ Lenton joined08:52
← Sound left08:53
→ JSharpe joined08:54
→ srji joined08:55
→ Sendoushi joined08:56
← msmto left08:57
→ xequence joined08:57
← intellix left08:58
→ n-st joined08:58
→ Willow joined08:59
→ Hounddog joined08:59
→ a_le joined09:01
→ whyy joined09:01
→ mladi joined09:02
→ Sound joined09:03
→ Gizburdt joined09:03
→ Advocation joined09:04
← Arkaniad left09:04
→ h0bbit joined09:04
→ intellix joined09:05
← a_le left09:05
← blorbx`` left09:06
→ blorbx`` joined09:06
→ crxz0193 joined09:09
← hoschi left09:10
→ morphix0 joined09:10
→ beautifulmind joined09:11
→ Not_Whatever joined09:12
← whyy left09:12
→ delight joined09:13
← alexbligh1 left09:13
→ alexbligh1 joined09:13
→ plitzenberger joined09:14
→ PaulCape_ joined09:14
← PaulCapestany left09:14
dkn64738_afk → dkn6473809:15
→ zumba_addict joined09:15
→ msmto joined09:15
← zumba_ad_ left09:16
→ zumba_ad_ joined09:16
← marduk191 left09:17
→ Arkaniad joined09:18
→ bgerber joined09:19
donnib when would -no-ff would be nice to use when pulling changes from remote ?09:19
→ nkouevda joined09:19
donnib i mean when is it useful ?09:19
← zumba_addict left09:20
osse it is useful when you want a merge commit even if the merge can be resolved as a fast-forward09:20
← bandroidx left09:20
osse which is something you want to do if you want to make it part of history that a merge in fact did happen09:20
donnib osse: and when is that particular useful ? I mean if i have an clone repo and made no changes why would i want that ?09:20
osse to have a certain structure to your history09:20
→ bandroidx joined09:21
osse in that case you wouldn't want that09:21
(well, you could but you would be a weirdo :P )09:21
donnib so i still don’t see the use of it09:22
i mean either i have changes and i need to show that changes have been merged or not09:22
i am trying to figure out when that would be useful09:22
osse it is useful in the same scenarios merge --no-ff is useful.09:23
← nkouevda left09:24
→ thevishy joined09:24
← Arkaniad left09:24
osse if you only use 'pull' to "get updated" so to speak then it's never useful09:24
→ jds joined09:25
jds Heya09:26
→ Arkaniad joined09:26
osse heeeeey ya09:26
(outkast)09:26
→ ciampix joined09:26
qsuscs fuuuuuuu09:27
now i got a, eh, when you always hear a song, earworm?09:27
osse I have that all the time09:28
← Gizburdt left09:28
osse I constantly get reminded of songs when people write stuff09:28
→ awkwardusername joined09:28
jds I'd like to make a branch point to my current head. Is there a recommended way of doing so, short of editing .git/refs/heads/foo ?09:29
grawity jds: git branch <branchname>09:29
osse jds: yes. 'git branch foo'09:29
jds Avoiding the "git checkout foo; git reset --hard HEAD@{1}" would be nice because it's quite a big checkout09:29
grawity uh09:29
jds osse, grawity: sorry, existing branch09:30
osse jds: yes. 'git branch -f foo'09:30
grawity jds: ah, see the --force option09:30
jds ah, handy09:30
thanks09:30
grawity I've no idea what you were talking about the checkout/reset thing, though.09:30
qsuscs git update-ref ftw!09:30
← RawProduce left09:30
→ ssarah joined09:31
→ masood joined09:32
← Arkaniad left09:32
→ Gizburdt joined09:33
← Tuna-Fish left09:33
→ Tuna-Fish joined09:34
← knighthawk left09:34
→ knighthawk joined09:34
→ whyy joined09:35
← masood left09:36
→ ChaosV joined09:36
← ChaosV left09:37
→ ilogger2 joined09:44
→ ClarusCogitatio joined09:44
→ _hockey567 joined09:44
→ JSharpe joined09:44
→ monty joined09:44
→ vifon joined09:44
→ bezet joined09:44
→ digitalextremist joined09:44
→ Polymorphism joined09:44
→ _5kg joined09:44
→ sokoll joined09:45
→ deciph3r joined09:45
→ Sendoushi joined09:45
→ Alenah joined09:45
→ Seveas joined09:45
→ edwardly joined09:45
qsuscs did its path stay the same in history?09:45
→ weltall joined09:45
→ synfinatic joined09:45
→ silverf0x joined09:45
→ hoschi joined09:45
→ bernalex joined09:45
→ linduxed joined09:45
→ kapu joined09:45
→ Guest12405 joined09:45
spacediver yes09:45
originally git log returned 2 commits09:45
addition, deletion09:45
→ terribleplan joined09:45
spacediver filter-branch rewritten stuff from addition up09:45
→ stealthii joined09:45
qsuscs is the addition commit reachable?09:46
→ midnightmagic joined09:46
aidalgol Is there some command to grep git commits (i.e. the changesets)?09:46
→ the8thbit joined09:46
→ Friss joined09:46
→ hyperair joined09:46
qsuscs aidalgol: the commit diff content or the message?09:46
aidalgol the diffs09:46
→ meatmanek joined09:47
→ blakbunnie27 joined09:48
→ eletuchy_ joined09:49
→ calculon joined09:49
→ rs0 joined09:49
→ Tuna-Fish joined09:49
→ blarson joined09:49
→ fstd_ joined09:50
spacediver it was reachable before i pushed rewritten branch09:50
in the new repo it is not found09:51
→ delight joined09:51
spacediver in original repo I have 2 sets of commits — original and rewritten09:51
→ jonmasters_ joined09:51
spacediver rewritten repo has the same commits before addition, after that all new commits09:52
which i consider correct rewrite09:52
→ nnyk_ joined09:52
→ Panicky joined09:52
spacediver git show blob shows blob, which means that it WAS pushed along with the branch. how to see why exactly?09:53
qsuscs are you sure that you didn’t rebase onto the addition commit and the file is now present in HEAD?09:53
→ Arkaniad joined09:53
← eletuchy_ left09:53
spacediver now to check that? I've only filtered branch and pushed branch, no rebase09:54
why no commits are linked to blob?09:54
→ rgr joined09:54
fstd_ → fstd09:54
→ harish_ joined09:54
qsuscs well, is the file present in HEAD?09:55
← bernalex left09:55
→ bernalex joined09:55
→ hkon_ joined09:55
→ leeN joined09:55
→ vtunka joined09:56
→ c74d joined09:57
spacediver how to check that?09:58
sorry for silly q:)09:58
qsuscs ls path/to/file09:58
spacediver nope09:58
qsuscs i’m sure there is a way to check, but i don’t know, you could post to the !list09:59
gitinfo [!official] Some official resources for git. Homepage: http://git-scm.com/ | Source repo: https://www.kernel.org/pub/scm/git/git.git/ (Mirrrored: https://github.com/git/git/) | IRC channel: You're in it. | Mailing list: http://vger.kernel.org/vger-lists.html#git09:59
→ msmto joined10:00
spacediver ok, thanks,10:01
→ zumba_addict joined10:01
aidalgol git-log -S does the job.10:03
→ achadwick joined10:05
Polymorphism → SittingShiva10:08
→ martinklepsch joined10:08
→ abra joined10:11
→ kcj joined10:12
→ msankhala joined10:12
→ TimmyTones joined10:13
→ Corey84 joined10:16
→ plitzenberger joined10:17
→ barneybook_8 joined10:18
→ jophish joined10:18
→ stamina joined10:18
→ xequence joined10:19
→ ValicekB joined10:19
← Corey84 left10:20
→ Corey84 joined10:21
← Corey84 left10:22
← Tuna-Fish left10:22
→ Corey84 joined10:22
→ nfk joined10:24
→ Gizburdt joined10:24
→ KevinTS joined10:26
→ evenflow joined10:26
← hyperair left10:27
→ z8z joined10:28
← c74d left10:28
→ seschwar joined10:28
→ Tuna-Fish joined10:28
← jophish left10:29
→ c74d joined10:31
← hoschi left10:31
← Corey84 left10:31
→ intellix joined10:31
→ Corey84 joined10:31
← martinklepsch left10:33
← Panicky left10:34
→ m0nkmaster joined10:34
← Gizburdt left10:35
→ whyy joined10:35
→ martinklepsch joined10:36
← martinklepsch left10:37
→ Willow joined10:40
← stamina left10:41
← bezet left10:42
← Corey84 left10:43
← TimmyTones left10:44
→ EminenceHC joined10:45
→ kevinb joined10:45
← Willow left10:45
→ Corey84 joined10:46
→ t4nk111 joined10:46
→ skylite joined10:47
→ martinklepsch joined10:48
→ m0viefreak joined10:49
← intellix left10:49
← EminenceHC left10:50
→ morphix0 joined10:51
← kcj left10:52
→ Gizburdt joined10:52
→ bezet joined10:53
→ diegovio1 joined10:54
← Corey84 left10:54
diegovio1 → diegoviola10:55
→ el_fuego joined10:55
→ daito joined10:55
← t4nk111 left10:55
→ Corey84 joined10:55
→ netj joined10:55
→ Willow joined10:56
→ arup_r joined10:58
← arup_r left10:58
→ BSaboia joined10:59
→ arup_r joined10:59
→ srji joined11:01
← Arkaniad left11:02
→ eletuchy joined11:03
→ Reskp joined11:04
← whyy left11:04
← xequence left11:05
→ Arkaniad joined11:06
← barneybook_8 left11:08
← daito left11:08
← martinklepsch left11:09
→ asakura1 joined11:09
→ beautifulmind joined11:09
→ joki joined11:10
← beautifulmind left11:11
→ beautifulmind joined11:11
← Tuna-Fish left11:12
← beautifulmind left11:13
→ beautifulmind joined11:13
→ HanClinto joined11:13
→ SethM joined11:15
← z8z left11:16
→ listochkin joined11:18
→ whyy joined11:19
← SethM left11:19
→ z8z joined11:19
→ nkouevda joined11:21
→ freestyl3r joined11:21
← plitzenberger left11:22
→ Tuna-Fish joined11:23
→ HanClinto_ joined11:24
← m0nkmaster left11:25
← nkouevda left11:26
→ ferreum joined11:27
← HanClinto left11:28
→ ngrt joined11:29
← HanClinto_ left11:29
→ groolot joined11:29
← groolot left11:29
→ diegovio1 joined11:31
→ nuck joined11:31
→ plitzenberger joined11:32
diegovio1 → diegovio1a11:32
← diegoviola left11:35
← freestyl3r left11:35
→ wild_oscar joined11:36
→ nkambo joined11:36
→ freestyl3r joined11:36
→ TimmyTones joined11:36
wild_oscar does anyone have experience with egit on Eclipse. I'm finding it strange that for a multi-module maven project egit is always adding the modules to the parent project's .gitignore file11:37
→ Advocation joined11:38
→ Mega joined11:40
← srji left11:43
diegovio1a → diegoviola11:43
← Tuna-Fish left11:43
→ Tuna-Fish joined11:44
→ knownasilya joined11:45
← Mega left11:45
← Gizburdt left11:45
→ Mega joined11:46
← Mega left11:46
→ intellix joined11:46
→ maveneagle joined11:48
→ Mega joined11:48
← plitzenberger left11:49
→ Gizburdt joined11:49
← Mega left11:49
→ xobb joined11:51
→ jottr joined11:51
→ goTAN joined11:51
→ DoniRS joined11:52
→ srji joined11:52
→ svector joined11:53
← svector left11:53
← Gizburdt left11:54
→ svector joined11:54
→ esclear joined11:54
← svector left11:54
→ svector joined11:55
← svector left11:55
→ vurma joined11:56
→ awkwardusername joined11:56
→ svector joined11:56
← svector left11:57
→ b_jonas joined11:57
→ svector joined11:57
← jottr left11:58
← svector left11:58
→ dotpot joined11:58
→ svector joined11:59
← svector left11:59
→ svector joined12:00
← svector left12:01
→ lsiv568 joined12:02
← dotpot left12:03
→ Mega joined12:03
← Mega left12:03
→ Mega joined12:03
→ carandraug joined12:03
→ skorgon joined12:04
← Arkaniad left12:04
→ plitzenberger joined12:04
← Mega left12:05
→ Arkaniad joined12:06
→ Mega joined12:06
← Mega left12:07
← Advocation left12:08
→ pecanha joined12:11
→ LBV_User joined12:12
→ gyyrog joined12:14
gyyrog hello12:14
gitinfo gyyrog: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.12:14
pecanha Hello! We're starting to use git here to develop an application. So, we created a remote(server) bare repository with the initial files all developers should have. Then, on each desktop, we cloned it to local machine, did few changes in branches, merge with our local master branch and then give a git push to remote(bare) server. Other desktops could pull back modified contents from other developers fine via remote/bare server. However, we can't12:14
see those changes applied to the bare repository, so what I need to do to update it with the changes sent from developers?12:14
← Reskp left12:15
→ mkoskar joined12:15
→ Reskp joined12:15
← Reskp left12:16
→ toedwy joined12:16
grawity pecanha: by "can't see those changes", what exactly do you mean? that the checked-out files aren't getting updated?12:18
pecanha grawity: I have ftp/ssh access to the remote repository and the files presented there doesn't contains the modified files.12:19
grawity hm12:19
pecanha grawity: however, if I pull from it, I can see the modified files12:19
grawity pecanha: bare repos normally don't have files checked out *at all*12:19
that's what "bare" means in the first place, just the repository database without any files 'extracted'12:19
gyyrog I want to use Git to track changes to netlogon scripts on a Windows domain controller. Anyone do this before, or have any suggestions? Would Git interfere anyway with replication?12:20
pecanha grawity: so, it is expected that the bare always have the original files ?12:20
grawity pecanha: no12:20
→ saorge joined12:20
grawity pecanha: like I just said, bare repos won't have any visible files at all12:20
pecanha: so either your repo isn't actually a bare one, or was incompletely changed to 'bare' in the middle12:21
pecanha grawity: I have that folder with files, and then I did a git --bare init12:21
grawity so it doesn't really *care* about those files then12:21
pecanha grawity: I could remove them and bare will continue to work fine. Is that correct?12:22
grawity right12:22
an 'init' doesn't automatically import anything, it merely creates the "database" for the repository12:23
← arup_r left12:23
pecanha grawity: so, what if I need a "central" repository and would like to open it via webserver online (like a production system) too? Which would be the best approach?12:23
→ Gizburdt joined12:23
grawity pecanha: see !deploy for a few methods12:23
gitinfo pecanha: 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.html12:23
lb1a pecanha, !bare12:23
gitinfo pecanha: an explanation of bare and non-bare repositories (and why pushing to a non-bare one causes problems) can be found here: http://bare-vs-nonbare.gitrecipes.de/12:23
→ arup_r joined12:24
pecanha In usual ways, I could send to remote via FTP/rsync when I'm ready to production, right?12:24
→ rypeck joined12:26
→ jophish joined12:26
→ WakiMiko joined12:27
→ cannonball joined12:28
← arup_r left12:28
→ simpleuser joined12:28
→ pnsrc joined12:29
→ Technodrome joined12:29
Technodrome i added a git tag to my local repo12:29
how do i push that tag to the remote repo?12:29
simpleuser Hello there. Let’s say I created a project, I initiated git. I created a first commit (to be able to return to the base project if needed). Then I create a folder and some files in it, I add a new commit.12:30
osse Technodrome: git push origin <tag>12:30
Technodrome: or git push --tags12:31
Technodrome yeah i just did a git push —tags12:31
simpleuser Then if I do a git reset --hard HEAD^ I should retrieve my project without the new folder, right?12:31
Technodrome it seemed to work12:31
→ godgodgodgo joined12:31
→ Skapadapaduu joined12:31
grawity pecanha: sure, though Git itself might be more efficient12:32
pecanha grawity: nice, thanks12:33
grawity have a non-bare clone at /var/www, add a hook in the bare repo to propagate updates to the clone, like that page suggests12:33
generally works well12:33
← toedwy left12:34
→ ORiON-- joined12:34
→ punchagan joined12:36
← Corey84 left12:37
← plitzenberger left12:37
→ bitkiller joined12:39
← Sendoushi left12:39
→ y_nk joined12:39
← y_nk left12:39
→ y_nk joined12:39
← y_nk left12:39
→ y_nk joined12:40
→ notjohn joined12:40
← srji left12:40
← pnsrc left12:41
← mkoskar left12:42
← el_fuego left12:42
→ thevishy joined12:44
→ aartist joined12:44
→ h0bbit joined12:44
pecanha grawity: whats a hook?12:45
grawity a script that runs after a certain event12:45
in this case, after something has been pushed into the bare repo (post-receive in `man githooks`)12:45
gitinfo the githooks manpage is available at http://jk.gs/githooks.html12:45
→ flaguy48 joined12:45
grawity see !deploy, the first and maybe 2nd methods12:45
gitinfo 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.html12:45
← flaguy48 left12:46
y_nk @Esya !12:46
→ EminenceHC joined12:46
Esya Yep?12:46
→ t-ask joined12:47
→ fr33domlover joined12:49
simpleuser Let’s say I created a project, I initiated git. I created a first commit (to be able to return to the base project if needed). Then I create a folder and some files in it, I add a new commit.12:49
Then if I do a git reset --hard HEAD^ I should retrieve my project without the new folder, right?12:49
→ Noldorin joined12:49
jast simpleuser: yep... destroying any uncommitted changes you might have.12:49
grawity note the difference between `git reset` and `git checkout`12:50
simpleuser jast: Then why my folders and files are still here?12:50
← aartist left12:50
jast maybe you didn't 'git add' the new folder before you tried to commit?12:50
_ikke_ Did you add those files before you comitted?12:50
→ srji joined12:50
moritz use 'git clean' to remove untracked files12:50
→ macabre joined12:50
simpleuser Oh, so I have to add them and then git reset ?12:50
← EminenceHC left12:50
simpleuser (it would explain a lot)12:51
← ORiON-- left12:51
jast no12:51
if you didn't add them before committing, the second commit doesn't actually contain them12:51
_ikke_ add them, and theb commit12:51
jast so, git never considers them worth looking at, and 'git reset' doesn't touch them either12:51
osse committing for the sole surposes of using reset HEAD^ sounds like a bad way of doing thigns12:51
→ mathieuhays joined12:51
t-ask Hi, I'm fairly new to development, while I already experimented a bit with git, I now want to get any github project and start modifying it's source. Yet, I don't know which is the recommended workflow to work on foreignprojects with git. Is there any good "top-down" tutorial for learning contributing code to existing projects?12:51
jast (you can use git log --name-status to see what happened in your previous commits)12:51
→ mizu_no_oto joined12:52
simpleuser Ok thanks12:52
→ ORiON-- joined12:52
jast t-ask: on github it usually works like this: 1) click the 'fork' button. clone your new repository. make changes, commit, push. do a pull request (another button on github).12:52
dang, I forgot to count up :)12:52
qsuscs pecanha: i use !cgit as a web interface, maybe i’ll switch to gitweb12:53
gitinfo pecanha: http://git.zx2c4.com/cgit/about/ web interface (cgi) for git repositories, written in C12:53
jast later on, when you update your fork with new changes from the original repository, things get a little more interesting, but that's the basics12:53
pecanha thanks12:54
→ lasers joined12:55
← z8z left12:55
t-ask I don't want to learn git commands itself, I just want to learn the "real" code contribution workflow. There are so many tutorials out there, which are boring by just teaching git commands like "Hello World" examples. What I search is something which directly digs into big projects and let me look over the shoulder of a real developer which he's following his/her daily git projects (not the github worklflow).12:55
← vtunka left12:56
t-ask /s/which/while12:56
→ Nizumzen joined12:57
moritz t-ask: the high level approach is: use pull requests12:58
donnib if i have a branch named A with 4 commits and i want the last two commits moved to a new branch named B how would i do that ? rebase onto and then reset of branch A ?12:58
→ Trudko joined12:58
moritz donnib: create branch B, do an interactive rebase of the last 4 commits where you throw out the two that you don't want12:58
donnib: then go to branch A, and git reset --hard HEAD~212:59
at least that's how I'd do it, because I tend to mess up rebase --onto12:59
← ORiON-- left12:59
→ HanClinto_ joined12:59
← bezet left13:00
donnib moritz: where does one start the interactive rebase from ? I assume you do that from the branch B ?13:00
osse alternative: create and checkout branch B at the point you want it to start, then cherry-pick A~1 A13:00
→ Sceorem joined13:00
donnib osse: but with cherry pick i would have to make the commits again right ?13:01
→ VictorBjelkholm joined13:01
moritz donnib: from HEAD~413:01
→ dhardison joined13:01
moritz donnib: and yes, branch B13:01
→ y_nk- joined13:01
t-ask moritz: sure, yet I'm searching a source where I can follow such an daily git workflow of someone who picks up an existing git project, modifies it, bekcome a new contributor and getting patches upstream. Someone who just explains what's the typical things to know eg.13:01
osse donnib: no, that's exactly what you do not have to do. that's the point of cherry-pick13:01
→ jerius joined13:01
← y_nk left13:01
dhardison can i pull another branch into a new directory in the same project?13:01
like a submodule i suppose?13:01
← zumba_addict left13:02
donnib osse: i thought cherry pick just take the changes and puts them in my working copy then i have to commit them on the branch B13:02
wild_oscar I'm finding it strange that for a multi-module maven project egit (on Eclipse) is always adding the modules to the parent project's .gitignore file. do you have any idea why Egit is behaving like this?13:02
qsuscs Seveas: does git-hub support adding refs/pull/…? if not, is it reasonable to add it?13:02
muep t-ask: I think it depends quite a lot on the project in question13:02
→ ORiON-- joined13:02
→ zumba_addict joined13:02
muep t-ask: e.g. some project might expect you to just mail them output of git format-patch. others might expect some kinds of pull requests13:03
osse donnib: you thought wrong13:03
← y_nk- left13:03
donnib osse: yes i understand, i must read upon cherry-pick13:03
→ vtunka joined13:03
→ mmitchel_ joined13:03
osse donnib: it can behave like that if you want to by adding --no-commit13:04
← jerius left13:04
t-ask muep: so there are no "general ruels" how to work with git with FOSS projects, eg? I mean, there must be typical steps and things you've learned to become a contributor. Things to consider and to know in general, steps to take eg.?!13:04
muep t-ask: also it depends on what kind of work you are doing. if it is just a simple-ish change, you might first make a commit that changes the thing you want and later rebase it until you feel you want to send it to the project13:04
→ y_nk joined13:04
→ Embassy joined13:04
→ nisseni joined13:04
→ arup_r joined13:05
→ kanupatar joined13:05
nisseni I've got a file in gitignore that's still being tracked. Do I have to do something special after I modified the gitignore file?13:05
kanupatar hi guys13:05
I am planning to host a git server in my laptop13:05
→ lb1a joined13:05
t-ask git server?13:05
kanupatar I was using SVN and planning to migrate to git13:06
t-ask: yes13:06
muep likely a server that hosts git repos13:06
osse nisseni: git rm --cached: gti commit13:06
arand__ dhardison: You might be able to use the !subtree merge strategy or the !git-subtree tool13:06
gitinfo dhardison: The subtree merge method is great for incorporating a subsidiary git repo into your current one with "unified" history. Read http://git-scm.com/book/en/Git-Tools-Subtree-Merging for more info, see also !git-subtree and !git-stich-repo.13:06
dhardison: git-subtree allows a subproject repository to be incorporated into a subdirectory of a main repository, and for the history of a subdirectory to be (re-)exported with reproducible results, making it possible to synchronise a repository with a subdirectory of another repo, see https://github.com/git/git/blob/master/contrib/subtree/git-subtree.txt for more info.13:06
→ qwyeth joined13:06
← zumba_addict left13:06
← Sceorem left13:06
muep t-ask: some more specific question might be easier to answer13:07
qsuscs kanupatar: there is no real “git server”13:07
muep t-ask: but writing a set of articles that'd give you all the common workflows would be a large-ish task13:08
→ plitzenberger joined13:08
qsuscs kanupatar: there are some different hosting solutions. the easiest, or from a different POV most complacited one is !gitolite13:08
gitinfo kanupatar: 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/13:08
← srji left13:08
→ m0ltar joined13:08
→ Mitchbones joined13:08
← Arkaniad left13:09
← arup_r left13:10
kanupatar qsuscs: thanks...how can I study the basics of git?13:10
qsuscs: is it possible to host git in windows machine?13:10
ToxicFrog kanupatar: basics as in basic usage, or as in basics of how it works internally?13:10
osse kanupatar: "basics" you say. The basics on how to use it or the basics on how it works?13:10
heh13:10
t-ask muep: well I try. While it's not easy to ask a specific thing about a subject which I currently try to learn. Therefore, I try to explain what I think of. I think of a virtual tutorial (video) which is similar the real world situation that I meet someone in person and he teaches me how he works with git and which things I have to consider.13:10
→ arup_r joined13:10
ToxicFrog high-fives osse13:10
qsuscs kanupatar: for either of what osse mentioned, i recommend the !book13:10
gitinfo kanupatar: 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 !parable13:10
osse I should change nicks to VenomousToad13:11
qsuscs kanupatar: on windows… it is possible, but might be a bit painful (easiest way might be !gitblit)13:11
gitinfo kanupatar: Gitblit is an open-source, multiplatform Git server written in Java, available both as a standalone and as a WAR file to be ran in other Java web servers. Though it's very easy to set up, it offers many features such as a neat web UI, sophisticated authentication options including LDAP, Groovy script support, a GUI-based management tool and upcoming, WIP features such as "tickets" (similar to pull requests).13:11
ToxicFrog kanupatar: it is possible, but like many things on windows, unpleasant13:11
→ zivester joined13:11
qsuscs (and, of course, “Saying that Java is good because it works on all platforms is like saying anal sex is good because it works on all sexes.”)13:11
ToxicFrog If you just want it for personal use (i.e. no need for access control or the like), easiest is probably just to install cygwin with the sshd and git packages and call it a day.13:12
→ sgambino joined13:12
qsuscs errr13:12
kanupatar: do you just want to develop locally or do you want to share with other computers?13:12
t-ask muep: I don't search for a "full" git tutorial, just something basic to start. I just want to (try to) contribute changes to a project, get more fun in doing this by just knowing very basic git commands. Things which happen often while developing with git. I don't care about the "special advanced" git commands. Just want to contribute source and get things done.13:14
← ORiON-- left13:15
← arup_r left13:15
→ arup_r joined13:15
muep t-ask: the first thing to do would be to get a clone of some project and make some new commits13:16
qsuscs t-ask: https://try.github.io is not too bad13:16
← nkambo left13:16
t-ask Sitting next to someone who uses git and just explains in ordinary words, what he's doing and where are the snares in daily git work13:16
→ masood joined13:16
→ ntio joined13:16
← arup_r left13:18
→ Asiajey joined13:18
→ tummy joined13:19
muep t-ask: basically with git, you can first just make some kinds of commits. then before publishing them, you can easily improve them if you notice that doing things differently would make the commits more suitable for their purpose13:20
← lb1a left13:20
← Tuna-Fish left13:20
→ ORiON-- joined13:20
→ Peuc joined13:20
→ daedeloth joined13:20
← punchagan left13:21
← lsiv568 left13:21
→ Sceorem joined13:21
→ lsiv568 joined13:22
→ tvw joined13:23
→ nkouevda joined13:23
→ _Vi joined13:23
donnib moritz: i tried your suggestion but somehow i am not doing it right becasue HEAD no points where the branch B has been created from which is the point where branch A was created from so saying HEAD~4 is probably wrong13:24
→ hcc23 joined13:24
→ enki joined13:25
donnib i did git rebase -i HEAD~4 on branch B which is pointing to 4 commits before branch A13:26
→ Corey84 joined13:26
← lsiv568 left13:26
donnib and what i want is to take the last two commits of branch A and put them so they looked as they were commited in branch B13:26
moritz donnib: and that's what you get if you remove the branch that are on A but shouldn't be on B13:27
← nkouevda left13:27
← Gizburdt left13:27
moritz (that is, remove the "pick ..." lines in the interactive rebase)13:27
→ Gizburdt joined13:27
→ gargola_ joined13:27
t-ask Sure, it's maybe a bit difficult to explain for me without being a native speaker. I think of a real world situation. I come to a developers workplace, he's already working at a huge project and I have just basic knowledge of git. Then we sit next to each other and I just look over his shoulder for one week or so. Just examining what's he doing and then slowly get the "git workflow" and things to consider while contributing to ordinary13:28
→ FortySix2 joined13:28
← msmto left13:28
→ it0a joined13:28
moritz t-ask: sounds like you should be looking for a developer in your town13:29
t-ask: try some local user groups maybe?13:29
donnib moritz: hmm but the pick lines showing up is showing commits not done in branch A but before creating that one13:29
→ knighthawk joined13:29
→ jedix joined13:29
→ mennea joined13:29
moritz donnib: something's wrong then13:29
→ berzerka joined13:29
moritz donnib: you said there were 4 commits on branch A, and doing a git rebase -i HEAD~4 shows you the last four commits13:29
← notjohn left13:29
moritz (unless there are merges inbetween)13:29
berzerka hi. how can i list the remote tracking branch of my local branches?13:30
← Technodrome left13:30
jedix hey, git in linux won't open my editor to create a commit message, it says "Aborting commit due to empty commit message."13:30
→ anEffingChamp joined13:30
jedix it should be opening vim13:30
..I think13:30
→ Arkaniad joined13:30
← pecanha left13:30
donnib moritz: i have done git rebase -i HEAD~4 while having the feature B checked out then i see 4 commits done prior to to the branch A creation13:30
jedix anyone know why this might be?13:30
→ hoschi joined13:30
moritz berzerka: git branch -vv should show it, otherwise some git-rev-parse @{u} with some options13:30
→ pothibo_ joined13:30
moritz jedix: try to set export EDITOR=vim13:31
berzerka jedix: did you do git config --global core.editor vim ?13:31
or that13:31
← evenflow left13:31
jedix $ env|grep vim13:31
VISUAL=vim13:31
EDITOR=vim13:31
VPAGER=vim -R -13:31
moritz donnib: then you didn't create branch B off of branch A13:31
→ Tuna-Fish joined13:31
jedix $ git config -l|grep vim13:31
user.editor=vim13:31
donnib moritz: true i didn’t13:31
moritz jedix: and is vim installed?13:31
donnib: you should have :-)13:31
berzerka moritz: thanks, -vv shows it.13:31
donnib morit13:31
← gargola_ left13:31
jedix yes, vim is installed13:31
core.editor?13:32
donnib moritz: yes i guess but if i did not want to base it on that ?13:32
→ lsiv568 joined13:32
t-ask moritz: well, possibly you are right. I just thought there is another well-known source somewhere in the web.13:32
donnib but another point from the master ?13:32
jedix core.editor does nto wokr either13:32
berzerka jedix: that's what the intro docs say: http://www.git-scm.com/book/en/Getting-Started-First-Time-Git-Setup13:32
moritz donnib: you can't use the second half of my solution without using the first half :-)13:32
→ Technodrome joined13:32
donnib both branch A and branch B starts from master13:32
but in different points13:32
aha13:32
moritz donnib: if you don't want to use my appraoch, start B off of master, and use cherry-pick13:32
→ dunpeal joined13:33
donnib moritz: i want but i am trying to figure out how to get it to work13:33
jedix I use user.editor, setting the core.editor didn't help13:33
moritz jedix: I'd try to use strace to find out if it tried to launch some editor13:34
← hoschi left13:34
jedix moritz: strace does not output any mention of vim13:34
← morphix0 left13:35
→ CKyle joined13:35
← whyy left13:36
dunpeal Hi. Is there a real problem with large commit messages? For instance, a 3kb commit message?13:36
grawity nope13:36
← Arkaniad left13:36
dhardison arand__: thank you13:36
dunpeal Basically there's a large bunch of test data related to the commit, and we'd like to add it - all the failues before an after the commit.13:36
dhardison i'm interested in learning about the different git "strategies" with -s where can i find a list of those?13:37
→ mkoskar joined13:37
moritz dhardison: man git merge13:37
gitinfo dhardison: the git-merge manpage is available at http://jk.gs/git-merge.html13:37
donnib moritz: if i’ve done what you said then if i want branch B to be based on another parent than branch A then i can rebase it on another commit from master right and achieve what i wanted e.g to base the branch B on a different point in master13:38
jedix this is git version 2.0.113:39
grawity dunpeal: there are several thousand commits in linux.git with messages over 4 kB, at least two over 12 kB (e.g. http://sprunge.us/PEGA)13:39
dunpeal grawity: thanks.13:39
→ Xethron joined13:39
← Xethron left13:39
→ Xethron joined13:39
grawity dunpeal: this shouldn't cause any performance issues unless you reach, I dunno, in the megabyte range?13:39
→ toedwy joined13:39
dunpeal We don't expect to surpass a MB size at this time, no.13:39
canton7 dunpeal, man git notes if you want a potential alternative13:40
gitinfo dunpeal: the git-notes manpage is available at http://jk.gs/git-notes.html13:40
← awkwardusername left13:40
grawity even then, megabyte is just a random number I came up with13:40
grawity still waiting for the commit-size script to finish13:40
canton7 (it may not be what you're after, and that's fine)13:40
moritz though notes aren't pushed/pulled by default, right?13:40
grawity right13:40
hmm, found two commits around 21 kB13:41
→ awkwardusername joined13:41
dunpeal Another question: we have a very large repository, and are concerned about download size and size-on-disk for users who are cloning it remotely. Would merging commits help with that?13:41
→ dragonball_ joined13:41
grawity merging in what way?13:41
dunpeal grawity: I.e. if we took 20 commites of size 10 MB each, and rebased (squashed) them to one 200 MB commit, would that make the size of the .git directory smaller?13:42
← CKyle left13:42
dunpeal And make cloning the repo faster?13:42
→ CKyle joined13:42
grawity depends; where are you getting the 10 MB from?13:42
canton7 20 * 10 = 200, so no :P13:42
qsuscs !binary13:42
gitinfo Storing binary files in git causes repo balloon, because they do not compress/diff well. In other words, each time you change a file the repo will grow by the size of the file. See !annex for some solutions13:42
grawity are you adding 10 MB of files every time?13:43
dunpeal these aren't binaries13:43
grawity or is it simply that you see .git growing by 10 MB every time?13:43
→ Dougie187 joined13:43
dunpeal grawity: it's a simplification, but basically we have a large codebase, that is currently built up from multiple commits13:43
← Peuc left13:43
qsuscs if you have a 10MB file and change some lines every commit, doing a git gc --aggressive every, say, every month should reduce the size a bit13:43
← eletuchy left13:43
→ derrzzaa joined13:43
← Corey84 left13:44
→ eletuchy joined13:44
← plitzenberger left13:44
dunpeal so, the total size of the codebase is about 2000 MB, and its history is "natural", in the sense that it is a chain of many small commits.13:44
grawity Git keeps full copies of every file, but eventually archives older objects to delta-compressed 'packfiles' (e.g. the clone download is one large packfile), so if all you're doing is editing text, then no, it wouldn't change much13:44
derrzzaa anyone recommend a tool for merging between two branches? in a scenario where there's no conflicts, and I can't just cherry-pick commits13:44
← beautifulmind left13:44
← vtunka left13:44
canton7 uh.. 'git merge'?13:45
Walex derrzzaa: 'git'? :-)13:45
grawity derrzzaa: is that total size just the .git repo or including the working tree?13:45
← kevinb left13:45
grawity derrzzaa: and with loose objects or without?13:45
derrzzaa kaleidoscope as a difftool kinda works, but doesn't let you add if they're added on one branch13:45
dunpeal grawity: thanks. in my case, we're actually mostly adding new files, not editing old ones much13:45
← esclear left13:45
derrzzaa huh? I'm doing a mix of things atm, checkout out files from the other branch if I need them13:45
grawity dunpeal: ah, in that case, you wouldn't see any noticeable change13:45
← pothibo_ left13:45
derrzzaa that's less useful if I only need some commits13:45
or hunks rather13:45
→ pothibo joined13:46
dunpeal grawity: I see, so it matters even less if we're only adding files13:46
→ Patrick_Baitman joined13:46
→ Peuc joined13:46
dunpeal grawity: so what the packfiles optimize is packing multiple changes to the same file13:46
like, I imagine they store deltas.13:46
grawity practically the only difference would be a few kB per commit (depending on how many directories are touched) and even that gets compressed eventually13:46
→ kevinb joined13:47
grawity yes, packfiles just delta-compress similar objects13:47
← h0bbit left13:47
jedix so, it is not happy with my user compiled vim.13:47
grawity not necessarily the same file; the compressor just looks for anything that would compress well13:47
→ ascarter joined13:47
← LBV_User left13:48
canton7 I don't think it tries to compress different sorts of object13:48
dunpeal grawity: ah, so it can detect content equivalence across files.13:48
grawity yes13:48
← dragonball_ left13:48
grawity that's how Git can get away with storing complete snapshots for each and every commit13:48
← eletuchy left13:48
← FortySix2 left13:48
← Xethron left13:49
→ srji joined13:49
→ gregsmith joined13:49
← srji left13:49
→ przemkovv joined13:49
→ srji joined13:49
← jedix left13:50
→ Ankhers joined13:50
→ rolledback joined13:50
grawity one of the suggestions for managing huge unchanging files is to have something like git-annex where the Git repo only tracks references (e.g. symlinks), and a helper tool downloads the actual files over HTTP or rsync on demand13:51
→ FortySix2 joined13:51
→ pweaver_ joined13:51
← CKyle left13:51
→ CKyle joined13:51
← hcc23 left13:51
→ hcc23 joined13:52
dunpeal that would be nice.13:52
→ kristofers joined13:52
→ Neozonz joined13:52
← Neozonz left13:52
→ Neozonz joined13:52
← jophish left13:52
dunpeal it would also be great to have some sort of resumable clone.13:52
← CKyle left13:52
← awkwardusername left13:52
dunpeal due to the size of the repo, we're getting complaints from users with slow connections.13:52
→ CKyle joined13:53
→ awkwardusername joined13:53
grawity it'd be nice but I don't think anybody tried to implement it13:53
→ Advocation joined13:53
→ Megaf joined13:53
← Ankhers left13:53
grawity it's possible to only download several latest commits using a shallow clone, with --depth=10 or something, and then download deeper history later13:53
dunpeal it takes them over an hour to clone the repo. since clone is an all-or-nothing process, any disruption in the connection during that time necessitates restarting from scrach13:54
grawity: thanks, I wasn't aware of --depth13:54
← Advocation left13:54
grawity but if you only *add* files and rarely change/remove them, then it sounds like the latest commit by itself would be just as large as the files themselves13:54
→ Ankhers joined13:54
dunpeal grawity: there was actually a GSoC project some time ago trying to implement resumable clones, but it was never followed through.13:54
grawity (meanwhile, --depth is for situations where the *history* weighs a lot)13:54
a resumable clone or git-annex would help, though.13:55
← rolledback left13:55
dunpeal grawity: I wonder if you can start from the root commit, and incrementally build up the repo from that.13:55
because that would fix the issue even for large repos like ours.13:56
→ rolledback joined13:56
← CKyle left13:56
← pweaver_ left13:56
dunpeal (in which each snapshot itself is very large)13:56
grawity from root, no, the protocol doesn't allow that13:56
→ CKyle joined13:56
← masood left13:57
grawity (not only because it takes a while to determine what the root commit _is_, and because there can be many of them, if several repositories get merged into one)13:57
← Nizumzen left13:57
grawity though that'd be interesting13:57
→ crxz0193 joined13:58
dunpeal grawity: yeah, you could always generalize to being able to clone a particular commit13:58
← Gizburdt left13:58
wild_oscar say I am working in a feature which is 2-4 days long. I create a new branch from my development branch and start working on it, committing locally from time to time. Now a colleague needs to work on something that needs this new feature. what's the best practice? should I update the development branch with the partially developed feature?13:58
← crxz0193 left13:58
dunpeal then we could just tell folks to clone from a particular hash.13:58
→ fatalhalt joined13:59
dunpeal wild_oscar: yes, what's the altenative?13:59
→ Gizburdt joined13:59
← Gizburdt left13:59
moritz wild_oscar: if it's ready, merge it into development; if not, push it as a separate branch13:59
→ Gizburdt joined13:59
wild_oscar dunpeal: the feature is not ready yet13:59
→ sohail_ joined13:59
dunpeal wild_oscar: another thing you may do, if other users of the dev branch don't wish to be impacted, is branch off from the dev branch, and merge into that branch14:00
i.e. create dev_with_partial_feature, which is a branch of dev with your feature branch merged in.14:00
→ Destos joined14:00
→ hj2014 joined14:00
wild_oscar dunpeal: and then ask the colleague to checkout that branch?14:01
dunpeal wild_oscar: yup.14:01
→ dogeydogey joined14:01
dunpeal you guys can work on that branch, while all the other devs stay on the dev branch.14:01
you can then incrementally merge in the dev branch and/or your feature branch as you see fit.14:02
wild_oscar I'm 2-3 weeks under my first real attempt to user git - forked a opensource project we're updating. the think I'm currently struggling with the most is...*branches, branches everywhere* - it seems I'm creating *a lot* of branches and that it soon will be really hard to see what's been merged, what hasn't been...14:03
→ ehmatthes joined14:03
← awkwardusername left14:03
← przemkovv left14:03
dunpeal wild_oscar: branching is pretty much the main reason to use Git, otherwise SVN is sufficient.14:03
canton7 !lol is useful for visualising14:03
gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all14:03
→ awkwardusername joined14:04
ToxicFrog wild_oscar: !gitk14:04
...huh, do we not have a trigger for that?14:04
canton7 what would it say?14:04
→ Gamecubic joined14:04
markmarkmark !gka14:04
gitinfo For a better way to view the reflog, try: gka() { gitk --all $(git log -g --format="%h" -50) "$@"; }; gka14:04
markmarkmark maybe?14:04
canton7 that's just the reflog :P14:04
dunpeal wild_oscar: one solution is to use a good graph visualization tool, like these guys are suggesting.14:04
qsuscs !visualize14:04
gitinfo You can visualize whole repo graph with gitk GUI viewer (gitk --all HEAD) or !lol for terminal view (git log --oneline --graph --decorate --all)14:04
dunpeal another is to learn `git log` really well, in which case you can reason about the repo using textual analysis, e.g. find merge bases and such.14:05
gyyrog Anyone see any issues with running Git in a windows netlogon directory?14:05
wild_oscar dunpeal: about 6 months ago when someone was mentioning that the problem was that *there was no good tool on linux* for visualisation14:06
qsuscs .trigger_edit gitk A tool to view commit history and the diffs they introduced, branching and merging, etc. It is shipped with git, run `gitk` from inside a repository.14:06
gitinfo qsuscs: Okay.14:06
→ LBV_User joined14:06
qsuscs any objections on that trigger?14:06
dunpeal wild_oscar: that's false. gitk is nice, and for me personally git log with --graph and decorations is good.14:06
I vaguely recall there are other visualization tools.14:07
ToxicFrog There are. I keep coming back to gitk because, despite being extremely ugly, it works the best,14:08
dunpeal wild_oscar: in my experience, usually the problem isn't the visualization tool, but inexperience with branching.14:08
← msankhala left14:09
qsuscs ToxicFrog: nothing is ugly again when you have seen java swing stuff14:09
dunpeal wild_oscar: if you master branching, any question you may have about branching states can be answered by a textual command.14:09
ToxicFrog qsuscs: the existence of Swing, while tragic, does not prevent other things from being ugly as well,.14:09
→ the_cybersphinx joined14:09
→ Advocation joined14:10
→ plitzenberger joined14:10
→ Arkaniad joined14:10
→ Hello71 joined14:10
← awkwardusername left14:10
→ joelrobichaud joined14:10
→ awkwardusername joined14:11
→ Corey84 joined14:12
→ diegovio1 joined14:12
diegoviola → Guest4201814:13
→ MblKiTA joined14:13
← zivester left14:14
← Tuna-Fish left14:15
→ Xiti` joined14:15
← the_cybersphinx left14:15
→ acharlieh joined14:16
← hkon_ left14:16
→ pat__ joined14:16
← Arkaniad left14:17
→ the_cybersphinx joined14:18
← Trudko left14:18
→ Relsak joined14:18
→ thiago joined14:18
→ Akiva-Thinkpad joined14:19
→ punchagan joined14:19
diegovio1 → diegovio1a14:20
→ joelm joined14:21
← lsiv568 left14:22
← pat__ left14:22
→ lsiv568 joined14:22
← srji left14:22
→ hyperair joined14:23
→ POIDHDGGFGFG joined14:23
→ nkouevda joined14:23
← lsiv568 left14:24
→ lsiv568 joined14:24
← POIDHDGGFGFG left14:24
→ int joined14:25
→ Tuna-Fish joined14:25
→ ujjain joined14:27
← dunpeal left14:27
→ pweaver_ joined14:27
← godgodgodgo left14:28
← nkouevda left14:28
→ epifanio joined14:29
→ msmto joined14:30
← msmto left14:30
← int left14:31
→ KevinTS_ joined14:31
→ supermat joined14:32
→ TDJACR joined14:32
← CKyle left14:33
← thiago left14:33
→ nyuszika7h joined14:33
nyuszika7h hi, how can I add an alias for something like 'git branch -D $1 && git push origin --delete $1'?14:34
→ hoschi joined14:34
→ EminenceHC joined14:34
→ skasturi joined14:34
→ msmto joined14:35
canton7 prepend with an exclamation mark - man git config has a note on it14:35
gitinfo the git-config manpage is available at http://jk.gs/git-config.html14:35
← KevinTS left14:35
→ FatDarrel joined14:35
← KevinTS_ left14:36
nyuszika7h do I just use $1 for the arguments?14:36
I know about the exclamation mark14:36
_ikke_ nyuszika7h: use a function14:36
← Advocation left14:36
nyuszika7h a function?14:36
Technodrome does a tag just point to the last commit?14:36
_ikke_ A tag points to the commit it was created on14:37
canton7 Technodrome, a tag poitns to the commit you told it to point to14:37
→ treehug88 joined14:37
→ polyzen joined14:38
Technodrome i just added a tag14:38
← nfk left14:38
Technodrome i didn’ give it a specific commit14:38
ah it points to the whole commit14:38
well i committed everything , then tagged, then pushed the tag to remote14:38
canton7 if you don't pass a commit, it will default to HEAD14:39
Technodrome so everything in that last commit has that tag? if so makes sense14:39
canton7 yeah, tag points to commit object14:39
← diegovio1a left14:39
← mkoskar left14:39
Technodrome there is 2 tags on the whole thing, 2 versions , and yeah i guess it just points to the thing i commited right before that14:39
or everything from the last tag commit?14:39
→ mkoskar joined14:39
canton7 "from the last tag commit"?14:40
← skylite left14:40
Technodrome sorrry if i’m explaing myself a bit wrong14:40
it just seems to add the tag to my last commit14:41
canton7 a tag is a standalone thing. a commit knows nothing of tags14:41
'git tag foo' is the same as 'git tag foo HEAD'14:41
i.e. create a tag that points to HEAD14:41
→ Xethron joined14:41
← Xethron left14:41
→ Xethron joined14:41
→ foist joined14:42
Technodrome yes14:42
i mean when you do a tag, its like a snapshot right of the branch?14:42
→ kpease_ joined14:42
Technodrome like i can always revert to this tag / release at any point14:42
← Akiva-Thinkpad left14:42
canton7 a tag is a thing which points to a commit14:42
Technodrome i never directed it to a commit14:43
canton7 a commit is a snapshot of your files at a point in time14:43
'git tag foo' is the same as 'git tag foo HEAD'14:43
i.e. create a tag that points to HEAD14:43
Technodrome a tag pointing to head points to what? sorry dont know much about HEAD14:43
← toedwy left14:43
Technodrome like a tag pointing to head just points to everything?14:43
→ alezandro joined14:43
canton7 HEAD point to the commit which is currently checked out14:43
*points14:43
Technodrome ah14:43
commit being the snapshot of the tree14:44
i kept thinking of commit as like a “file change” that got committed14:44
canton7 a commit has a committer, a date, and points to a snapshot of the tree14:44
nah, commits are snapshots14:44
Technodrome yeah got it14:44
canton7 read !bottonup14:44
Technodrome so when i tag something its pointing to the last commit?14:44
canton7 !bottomup14:44
gitinfo 'Git from the bottom up' starts with explaining the building blocks of git and proceeds to tell you how they fit together. http://ftp.newartisans.com/pub/git.from.bottom.up.pdf (http://ikke.info/git.from.bottom.up.pdf)14:44
canton7 so if you create a tag with 'git tag foo', it creates a tag pointing to the commit which is currently checked out14:45
→ ribasushi joined14:45
Technodrome i haven’t dug into git terminolgy and really how things really work, i just push roll and rollback eveyr now and then and now tag :P14:45
canton7 you can create a tag which points to something else: 'git tag HEAD^' or 'git tag mybranch' for example14:45
Technodrome yeah14:45
canton7 it's really, REALLY worth learning ;)14:45
seriously, 100x14:45
Technodrome i will14:45
after work i will spend a few hours reading on some more stuff14:45
← maveneagle left14:45
← TimmyTones left14:45
canton7 once you've got a basic grasp of the object model, everything else magically makes sense14:46
← FortySix2 left14:46
Technodrome with that being said, after i commited , tagged, and pushed the tag, the sysadmin pulled the release down and everthing works :P14:46
i had the wrong idea of what a commit was sadly enough14:46
canton7 sure - I didn't say nothing will work until you do some reading ;)14:46
hehe14:46
→ TimboTones joined14:46
Technodrome i though it was a “change” , not a complete snapshot14:46
canton7 other version control systems store them as diffs - mercurial for example14:47
and most of the time it's an implementation detail14:47
a commit also stores a reference to its parent - the commit on which it's based14:47
so git can figure out the diff easily enough14:47
← hcc23 left14:48
Technodrome yeah14:48
→ hcc23 joined14:48
→ Macaveli joined14:48
← pweaver_ left14:48
Technodrome so when i did a git push, doesn’t it literally send my complete snapshot over ? or does it just send the differences?14:49
how does that part work14:49
pasky_ it sends just the differences14:49
Technodrome but the actual commit is a snapshot though?14:49
pasky_ yes14:50
Technodrome so what are the differences called?14:50
canton7 there's a negotiation phase where it figures out what objects (commits, snapshots, tags, whatever) it has that the remote side doesn't have14:50
pasky_ sending just the differences is just an "optimization", not a fundamental thing14:50
Technodrome revisions? lol14:50
canton7 then it creates a packfile, which compresses all of those objects into a single compressed file14:50
then it sends the packfile14:50
Technodrome ah so its still a commit, it just figures out what to send14:50
to it doesn’t send the entire thing14:50
so*14:50
pasky_ the differences aren't first-class objects in git14:50
← thevishy left14:51
canton7 if you read git from the bottom up, you'll understand what objects are :P14:51
Technodrome is that a book?14:51
canton7 it sends all the objects which the remote side doesn't have, and which the remote side needs14:51
Technodrome ah14:51
nyuszika7h _ikke_: I get a syntax error if I include semicolons in the alias, escaped or not14:51
Technodrome sorry14:51
yah the pdf14:51
canton7 "thin packs" (the default now) also take into account what objects the remote side has when compressing - so it can say "object XXX is your object YYY plus *these*"14:51
that's all implementation, though14:51
Technodrome I will read that complete thing tonight14:51
← mizu_no_oto left14:51
Technodrome so no need for me to ask any more questions until that’s done :)14:51
→ ojacobson joined14:52
gitinfo set mode: +v14:52
→ thedodd joined14:52
→ morphix0 joined14:52
→ Juri_Agarin joined14:53
gitinfo set mode: +v14:53
→ Neozonz|Disc joined14:53
← Neozonz|Disc left14:53
→ Neozonz|Disc joined14:53
→ pat__ joined14:53
← pat__ left14:53
→ kriks joined14:54
Technodrome definitely appreciate it though canton7 and pasky_14:54
→ nickpresta joined14:54
→ Vivekananda_y510 joined14:54
canton7 :014:54
* :)14:54
→ FortySix2 joined14:54
→ codeFiend joined14:55
← kevinb left14:55
← Gizburdt left14:55
→ Neozonz|Disc2 joined14:55
← mkoskar left14:55
Juri_Agarin Another good reference for learning about Git's internals: http://git-scm.com/book/en/Git-Internals-Git-Objects14:55
← Neozonz left14:56
Juri_Agarin Staring at those pictures for a while helped me a lot.14:56
← anEffingChamp left14:56
→ pat__ joined14:57
← pat__ left14:57
→ shinnya joined14:57
← Neozonz|Disc left14:58
canton7 yeah, the book is required reading14:58
→ Neozonz|Disc joined14:58
→ dreiss joined14:58
canton7 bottomup is a shorter intro to just the object stuff, and imo it's better written than that section of the book14:58
although !doc14:58
gitinfo A list of useful documentation of all kinds is here: http://git-scm.com/documentation -- or try the wiki at http://git.wiki.kernel.org/. Also try typing "!book" "!cs" "!bottomup" "!parable" "!best_practices" or "!vcbe" or "!designers" here in IRC. !book is probably the most helpful.14:58
canton7 lots of very good articles out there14:59
← plitzenberger left14:59
canton7 (and lots of blog rubbish)14:59
qsuscs (meanwhile, hg folks say that their book is outdated and contains a lot of then-good-now-bad advice)14:59
→ toedwy joined14:59
← berzerka left15:00
→ pat__ joined15:00
→ bezet joined15:00
← Patrick_Baitman left15:00
← ngrt left15:00
← pat__ left15:00
← FortySix2 left15:00
← ValicekB left15:01
→ wwalker joined15:01
← FatDarrel left15:01
→ listochk_ joined15:02
← pothibo left15:02
→ israelpattison joined15:02
→ interru joined15:02
Juri_Agarin qsuscs: I wonder what has changed during the years, turning good advice to bad?15:02
wwalker I just forked a dead project on github. I want to cherry pick the commits of others who have forked it into my fork. Anyone got a URL15:02
← Neozonz|Disc2 left15:02
wwalker or what to search for to see how to best do this?15:03
→ surfdue joined15:03
→ JanC joined15:04
← Neozonz|Disc left15:04
→ kevinb joined15:04
→ Neozonz joined15:04
← Neozonz left15:04
→ Neozonz joined15:04
grawity clone your fork to your computer, then fetch all other forks into it15:04
wwalker I'm thinking create a branch for each remote, pull into those and cherry pick from there. Is there an easier way?15:04
grawity probably by adding remotes15:04
← bezet left15:04
grawity "create branch", "pull into it", that's all unnecessary15:05
← listochkin left15:05
grawity just add a remote and fetch it, and you'll have the commits15:05
← mennea left15:05
grawity if it is a one-time job, you could even `git fetch` the URL directly15:05
wwalker grawity: it's a one time job :-) Thank you!15:06
→ Gizburdt joined15:06
→ anEffingChamp joined15:06
→ _antino joined15:07
canton7 if you fetch the url directly, you also need to specify a branch. that branch then gets fetched into FETCH_HEAD15:07
osse the network thingy on github will display exactly which commits that are present in any forks but not in the original repo, btw15:07
← kevinb left15:08
← Gamecubic left15:08
→ The-Compiler joined15:08
→ zapb joined15:08
→ Gamecubic joined15:08
The-Compiler Often I use git-blame, and then just see a commit where I moved some lines or so. Is there some easy way to say "give me the second-to-last commit where this line was changed"?15:09
qsuscs Juri_Agarin: *shrug* or maybe it was already bad back then and even worse now, i have no idea15:09
← codeFiend left15:09
→ ValicekB joined15:09
← Neozonz left15:09
→ tang^ joined15:10
→ Neozonz joined15:10
→ FortySix2 joined15:11
bremner The-Compiler: maybe "git gui blame" helps15:11
→ kevinb joined15:11
→ dccc joined15:11
→ mementomori joined15:12
mementomori hi15:12
→ robarelli joined15:13
← robarelli left15:13
→ robarelli joined15:13
→ thiago joined15:13
The-Compiler bremner: whoa, I usually don't like the gui much, but for blame it's nice :)15:13
← hcc23 left15:14
→ hcc23 joined15:14
← Neozonz left15:15
mementomori I have 2 branches: master and wip. Now I'm working on wip. I've found a bug I want to fix in master. I think I've to do these steps: 1) add to the index the missing file. 2) stash the changes 3) checkout master 4) fix the bug and commit changes in master 5) checkout wip 6) stash pop.15:16
is my plan correct?15:16
→ darkbasic joined15:16
bremner mementomori: ok, but you'll have to merge from master if you want those changes on wip15:16
The-Compiler mementomori: so in wip you added a new file which isn't in the index at all currently?15:17
← punchagan left15:17
mementomori bremner, this is sure15:17
→ punchagan joined15:17
mementomori The-Compiler, actually they are in wip's index15:17
← calculon left15:17
The-Compiler then just stash/checkout/etc15:18
mementomori The-Compiler, ready to be committed but I dont want to commit before the bug in master is closed15:18
The-Compiler stash also stashes non-staged changes15:18
→ montchr joined15:18
→ Akiva-Thinkpad joined15:18
← TimboTones left15:18
→ TimmyTones joined15:18
bremner mementomori: it sounds like you fear commitment for no obvious reason15:19
← Akiva-Thinkpad left15:19
mementomori bremner, I like to commit just clean code to the repo15:19
bremner shrugs15:19
→ Wiz_KeeD joined15:19
The-Compiler (argh, this reminds me again I always thought stash would do what reset --hard does, ending up with repos with dozens of stashed changes)15:19
bremner I'd rather clean up my commits later than try to understand all the subtleties of stash15:19
Juri_Agarin How about instead of stashing, do a WIP-commit to your branch, fix stuff in master, rebase the wip-branch on top of the new master, and then squash/amend when you continue working on the branch-stuff.15:20
mementomori after the commit in master (and before the checkout of wip) what if I stash pop?15:20
The-Compiler what subtleties does stash have?15:20
→ kotk_ joined15:20
bremner The-Compiler: if it has no subtleties, why are we talking about it? ;)15:20
If I have to stop and think, making a commit is a better idea.15:21
→ maveneagle joined15:21
mementomori Juri_Agarin, this is a path but since I've never been in that scenario I'd like to try stash ;)15:21
← maveneagle left15:21
→ Gege joined15:21
→ xudongz joined15:21
← Gizburdt left15:22
← awkwardusername left15:22
→ awkwardusername joined15:22
← Guest42018 left15:22
→ Gizburdt joined15:22
mementomori anyway. If I stash pop while in master will stash switch to wip branch?15:23
→ vsgill joined15:23
The-Compiler no15:23
stash pop will pop on whatever branch you are on15:23
← gregsmith left15:23
→ gregsmith joined15:23
mementomori The-Compiler, ah... How can I know to which branch a stash belongs to?15:23
← MblKiTA left15:23
→ nkouevda joined15:24
→ tranzfuse joined15:24
mementomori or better. How one could know which branch was used to create a stash15:25
← ascarter left15:25
→ ascarter joined15:25
→ Pyrogerg joined15:25
← ascarter left15:25
→ freezevee joined15:26
→ MblKiTA joined15:26
→ ascarter joined15:26
→ pikaren joined15:26
xender mementomori: Not branch, commit. Git log15:26
The-Compiler mementomori: git stash list or git log15:26
mathieuhays git blame for a file15:26
mementomori is thinking about making a 'dirty commit' before checking out master...15:26
→ nfk joined15:26
canton7 makes WIP commits all the time15:26
canton7 git commit --amend and git rebase -i are really useful for sorting those out15:27
Juri_Agarin I even have an alias for doing wip-commits. :-)15:27
→ vtunka joined15:27
canton7 yeah, I aliassed 'git commit --amend --no-edit' to 'git amend'15:28
mementomori ok.. let's commit wip, fix master and merge into wip15:28
it looks like a lot easier15:28
→ martinklepsch joined15:28
Pyrogerg I'm new to git and I have what's probably a very simple question. I want to get a tarball for a particular release version of some software hosted on bitbucket, e.g. https://bitbucket.org/lgautier/rpy2/commits/tag/RELEASE_2_4_215:28
← msmto left15:28
← martinklepsch left15:28
Pyrogerg How can I construct a URL to get that? It should be programatic, as I want to code this into a Makefile for a package in pkgsrc.15:29
→ martinklepsch joined15:29
Juri_Agarin mementomori: if your branch is local (not published to anyone else), you could even try your wings with git rebase after the master is fixed.15:29
xender mementomori: Merge? Use rebase for liner history's sake :P15:29
Or cherry-pick if it's only one commit15:30
Both boils down to same in that case15:30
Juri_Agarin xender: if the branch will be later merged to master, cherry-picking will lead to confusing history.15:30
← xudongz left15:30
bremner not if you rebase first15:30
mementomori xender, what's wrong with merge?15:30
→ cworth joined15:30
Juri_Agarin bremner: if you rebase first, you have nothing to cherry-pick anymore.15:30
xender mementomori: Merging of WIP-commit made on dirty master branch into cleaned-up version?15:31
← freezevee left15:31
→ xudongz joined15:31
xender mementomori: In that situation what will lead to confusing history is merge15:31
→ DeathIncarnate joined15:31
← ValicekB left15:31
xender !postproduction15:31
gitinfo So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.15:31
← derrzzaa left15:31
Juri_Agarin mementomori: generally merging leads to uninteresting history, that's harder to read, when you're e.g. debugging. Using git bisect will also become harder.15:31
xender Oh, also maybe !sausage and !perfect15:32
gitinfo [!sausage_making] Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.com/GitBestPractices/#sausage and !perfect15:32
[!postproduction] So, you want to make your commit history look pretty before pushing? http://sethrobertson.github.com/GitPostProduction talks you through how to use 'rebase -i' to do this.15:32
mementomori xender, nono... I'll merge master into wip and continue to work in wip after the bug is closed. master will be always clean15:32
xender Oh, third one is alias to first one. Good to know, I can never remember those redundant triggers15:32
mementomori: Why not just rebase wip onto new master?15:33
mementomori: As long as wip is not being published, rebasing is the way to go15:33
← rgr left15:34
mementomori xender, never heard about rebase before now. I've to see what it does15:34
← ntio left15:34
xender mementomori: And if you intend to publish it, it certainly shouldn't be just named "wip". It would become normal feature branch then and some descriptive name would be better15:34
← hoschi left15:35
→ ntio joined15:35
xender mementomori: There's chapter on it in the !book. man git-rebase may also be helpful in explaining it15:35
gitinfo mementomori: 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 !parable15:35
mementomori: the git-rebase manpage is available at http://jk.gs/git-rebase.html15:35
→ mark06 joined15:35
← mark06 left15:36
mementomori xender, I'll eventually merge wip into master when work is done. and delete wip. (I actually use another name for wip)15:36
xender Hey, why git diff -M -C somebranch -- somedirectory shows only deletions while those files are in my working tree?15:37
They are unstaged (not tracked), but manpage says that this form of git diff invocation compares commit with worktree15:37
mementomori: Why not rebase?15:37
Unless you need to publish wip branch in the meantime (like for collaborating on it), there's no reason to use a merge15:38
→ brownbathrobe joined15:38
← vurma left15:38
xender Not knowing how rebase works in not an excuse to refuse learning.15:38
← dreiss left15:39
mementomori xender, as I said I wasn't aware of rebase so my working flow was: checkout a new branch from master. do work. chechout master. merge the branch. delete the branch.15:39
xender It's not rocket science also15:39
← nisseni left15:39
→ ValicekB joined15:39
mementomori xender, I'll see rebase asap15:40
xender Hmm, ok, I can think of a reason to do that - prefering to have "real" history rather than edited one. But editing gives some possibilities15:41
← brownbathrobe left15:42
xender Also, hmm. I can even think of scenarion with rebase + non-ff merge - for that would make bit bisect easier, while retaining property of enclosing whole work in separate history line which can be skipped15:42
That could be useful in git-flow15:42
mementomori xender, I'll look like naive but I thought it was the way to work with git...15:42
← dhardison left15:43
xender mementomori: There's more than one way to do it ;)15:43
It's closer to being "the way" in Hg, but even there there is a rebase, as an extension15:43
← Macaveli left15:44
→ eletuchy joined15:45
← intellix left15:45
→ darkxploit joined15:45
→ pheaver joined15:45
→ skylite joined15:46
→ sonicparke|bradm joined15:47
→ intellix joined15:47
→ rgr joined15:49
← eletuchy left15:49
← DeathIncarnate left15:50
← listochk_ left15:50
→ bluesm joined15:50
← m0ltar left15:53
← dccc left15:53
→ exobit joined15:55
← Ankhers left15:55
→ iarna joined15:56
→ m0ltar joined15:56
→ Advocation joined15:56
← lsiv568 left15:57
→ lsiv568 joined15:57
→ plitzenberger joined15:58
← Wiz_KeeD left15:59
→ DeathIncarnate joined15:59
→ zumba_addict joined15:59
← nkouevda left16:01
→ devbug joined16:01
← anEffingChamp left16:02
→ a_le joined16:03
← Xethron left16:04
→ nkouevda_ joined16:05
→ Quevin joined16:05
→ eletuchy joined16:06
← nkouevda_ left16:07
→ netto_ joined16:07
← netto_ left16:07
← SittingShiva left16:09
→ lite_ joined16:10
← eletuchy left16:11
← martinklepsch left16:13
← plitzenberger left16:14
→ nerdfiles joined16:14
← lite_ left16:15
→ bhldev joined16:15
→ kballou joined16:15
→ Nizumzen joined16:15
→ Doyle joined16:16
→ cilly joined16:16
← cilly left16:16
Doyle Hello. Does GHE accept multiple LDAP targets for redundancy, or just one?16:17
→ cilly joined16:17
→ hkon_ joined16:17
→ freezevee joined16:17
← seschwar left16:17
← thedodd left16:17
← cilly left16:19
→ thedodd joined16:19
→ esclear joined16:20
→ smw joined16:20
→ mennea joined16:20
→ thevishy joined16:20
→ Ankhers joined16:20
← hkon_ left16:21
Juri_Agarin Doyle: some github-channel might have more knowledge about such.16:21
← Gizburdt left16:22
Doyle ah, true. Thanks Juri_Agarin16:22
← lsiv568 left16:22
→ lsiv568 joined16:23
iarna → iarna_16:23
← it0a left16:24
iarna_ → iarna16:24
→ Neozonz|Disc joined16:24
← Technodrome left16:24
← Neozonz|Disc left16:24
→ Neozonz|Disc joined16:24
→ sdrfnord1 joined16:24
← iarna left16:25
→ TimboTones joined16:25
→ hoschi joined16:25
→ iarna joined16:26
← TimmyTones left16:26
→ aantix joined16:26
→ stamina joined16:27
← lsiv568 left16:27
← vtunka left16:27
← kballou left16:28
← hcc23 left16:28
→ hcc23 joined16:28
← skorgon left16:28
→ skorgon joined16:29
→ TimmyTones joined16:29
← awkwardusername left16:30
→ tjj joined16:30
← hyperair left16:30
→ Panicky joined16:30
← DoniRS left16:31
← Sceorem left16:32
← a_le left16:32
→ Technodrome joined16:32
→ a_le joined16:32
← TimboTones left16:32
→ LucasTT joined16:32
→ codeFiend joined16:33
→ nkouevda joined16:33
← achadwick left16:34
→ teobo joined16:36
→ maveneagle joined16:36
← maveneagle left16:36
→ txt23 joined16:37
txt23 How do you roll back a couple of commits?16:38
milki txt23: !fixup16:38
gitinfo txt23: So you lost or broke something or need to otherwise find, fix, or delete commits? Look at http://sethrobertson.github.com/GitFixUm/ for full instructions, or !fixup_hints for the tl;dr. Warning: changing old commits will require you to !rewrite published history!16:38
← nkouevda left16:40
← codeFiend left16:40
→ jophish joined16:40
→ mikecmpbll joined16:41
→ it0a joined16:41
← txt23 left16:42
→ Kicer86 joined16:43
→ ciampix_ joined16:43
→ lasserix joined16:43
→ codeFiend joined16:44
← codeFiend left16:44
→ JDHawke joined16:44
→ SittingShiva joined16:45
→ Sijdesign joined16:45
Hello71 !revert16:45
gitinfo That's a rather ambiguous question... options: a) make a commit that "undoes" the effects of an earlier commit [man git-revert]; b) discard uncommitted changes in the working tree [git reset --hard]; c) undo committing [git reset --soft HEAD^]; d) restore staged versions of files [git checkout -p]; e) move the current branch to a different point(possibly losing commits)[git reset --hard $COMMIT]?16:45
← shinnya left16:45
→ codeFiend joined16:45
sjmikem isn't --soft implied?16:46
→ vurma joined16:46
Hello71 man git-reset16:47
gitinfo the git-reset manpage is available at http://jk.gs/git-reset.html16:47
← intellix left16:48
→ BBBThunda joined16:49
→ Xethron joined16:49
← Xethron left16:49
→ Xethron joined16:49
osse sjmikem: no, --mixed is16:49
jophish I did a shallow clone, and would like to actually have all changes for a longer history16:49
← codeFiend left16:49
→ Gizburdt joined16:50
→ jovy88 joined16:50
osse jophish: git fetch --unshallow16:50
← SittingShiva left16:51
jophish osse, ah I see. Thanks!16:51
← stamina left16:52
→ Polymorphism joined16:54
← Polymorphism left16:54
→ eletuchy joined16:54
→ Polymorphism joined16:54
← knighthawk left16:54
→ kballou joined16:58
← Relsak left16:58
← eletuchy left16:58
← Sijdesign left17:00
← Quevin left17:00
→ FatDarrel joined17:00
→ bezet joined17:00
← dogeydogey left17:00
← FortySix2 left17:02
← Xethron left17:02
→ lsiv568 joined17:02
→ bartman`` joined17:02
← FatDarrel left17:03
→ JeanCarloMachado joined17:03
← bitkiller left17:04
→ Xethron joined17:05
← bezet left17:05
← kpease_ left17:05
→ kpease joined17:05
→ skoop joined17:05
finster hey there. i have a file which was changed in, say, 5 of 100 commits in a givenrepository. can i somehow checkout these commits without going through git log, copying the respective hashes and doing a manuel git checkout <sha>?17:07
would be sufficient if i could check out the commits one after another17:08
← rgr left17:08
canton7 finster, what determines when to check out the next commit?17:09
← rolledback left17:09
→ tranzfus_ joined17:10
→ maestrojed joined17:10
osse finster: git rev-list --author=finster HEAD | while read sha; do git checkout $sha; your stuff here; done17:10
Eugene finster - man git-bisect17:10
gitinfo finster: the git-bisect manpage is available at http://jk.gs/git-bisect.html17:10
← JeanCarloMachado left17:10
← tranzfus_ left17:11
→ tranzfus_ joined17:12
finster i'll give it a shot. thanks to both of you17:13
← tranzfuse left17:13
→ alexbligh1 joined17:14
→ Relsak joined17:14
osse maybe:17:15
finster: git rev-list --author=finster HEAD | while read sha; do git checkout $sha; bash; done17:15
for each commit it starts a new one. then you can do whatever you want and when you're done you can exit the shell. Voila, new commit and a shell to do whatever you want and when you're done you can exit the shell. Voila17:16
→ pantsman_ joined17:16
→ msankhala joined17:16
→ SegFaultAX joined17:18
← hcc23 left17:18
→ hcc23 joined17:19
← TimmyTones left17:20
← gregsmith left17:21
← qwyeth left17:21
→ qwyeth_ joined17:21
← tjj left17:22
← msankhala left17:22
→ msankhala joined17:23
← msankhala left17:24
← Advocation left17:24
← vurma left17:24
→ glebihan joined17:26
→ parcs joined17:26
parcs hi17:27
← qwyeth_ left17:27
parcs why does git gc sometimes run twice in a row after running e.g. "git fetch" twice?17:27
→ pnsrc joined17:28
→ codeFiend joined17:28
→ traverseda joined17:30
← tvw left17:30
traverseda Is there any way to host a git repo on a CDN? I'm not really familiar with the git over http stuff17:30
canton7 you need something smart on the other end, if you're going over http (or almost anything else)17:31
→ Arkaniad joined17:31
→ ciampix__ joined17:31
← Xethron left17:32
traverseda canton7, so I couldn't link to specific files normally, and also have a sort of git "index" that git could clone from?17:32
Don't need to push to the CDN obviously.17:32
canton7 not sure what you mean17:32
← goTAN left17:32
← freestyl3r left17:33
← ciampix_ left17:33
← Arkaniad left17:34
traverseda Have the files in whatever is the latest head available as normal files hosted on a CDN. Also make the git history available.17:34
← hcc23 left17:35
→ hcc23 joined17:35
canton7 git can clone over http if there's something dumb on the other end ("dumb http"), provided that you've generated some required metadata files17:35
but you can't push over dumb http17:35
→ Arkaniad joined17:35
→ FortySix2 joined17:35
← thevishy left17:36
→ denisMone joined17:36
→ kpease_ joined17:36
← kpease left17:36
traverseda canton7, dumb http was the keyword I was looking for. Thanks.17:36
→ netto_ joined17:36
→ hyperair joined17:37
→ dreiss joined17:38
→ Mobutils joined17:39
← ValicekB left17:39
qsuscs but using dumb http is, well, dumb17:39
← FortySix2 left17:40
← BSaboia left17:40
→ qwyeth_ joined17:40
→ FortySix2 joined17:40
← ferreum left17:40
← TDJACR left17:41
→ Xethron joined17:41
← Xethron left17:41
→ Xethron joined17:41
→ TDJACR joined17:41
← kballou left17:41
← netto_ left17:41
← t-ask left17:43
ojacobson canton7: "well, kind of"17:44
← exobit left17:44
← kevinb left17:44
← silverf0x left17:44
← kapu left17:44
← synfinatic left17:44
← Alenah left17:44
← stealthii left17:44
ojacobson at least as of 1.7 or so git was able to push semi-dumbly over HTTP, for web servers with DAV enabled. :)17:44
→ Alenah- joined17:44
ojacobson But the implementation of it is so bad you never want to do it that awy17:44
→ synfinatic joined17:44
→ kevinb joined17:45
→ rolledback joined17:45
← dreiss left17:45
← ehmatthes left17:45
→ rgr joined17:47
canton7 I wouldn't quite call that dumb :P17:47
but yeah17:47
← zumba_addict left17:47
→ stamina joined17:48
→ zumba_addict joined17:48
→ kermit joined17:49
→ ValicekB joined17:49
→ Quevin joined17:50
→ akushner joined17:50
← DeathIncarnate left17:50
← parcs left17:51
← zumba_addict left17:52
→ vurma joined17:53
← netj left17:53
→ dccc joined17:53
→ netj joined17:53
kpease_ → kpease17:53
→ hahuang65 joined17:54
→ ropav joined17:54
← Nizumzen left17:55
→ eletuchy joined17:55
→ suffice joined17:55
→ diegoviola joined17:57
← pikaren left17:58
gyyrog Is it a bad idea to run Git in my domain controller netlogon folder? I want to have version control on my scripting folders17:58
?17:58
→ dogeydogey joined17:59
← kpease left17:59
→ lpax joined17:59
← Panicky left17:59
← eletuchy left17:59
← zapb left17:59
← Tuna-Fish left18:00
← akushner left18:00
→ Zamerick joined18:00
→ akushner joined18:00
→ pl joined18:01
→ dAnjou joined18:01
→ netto_ joined18:01
pl Hi. I have a stupid question, but I can't find the answer. I want to gitignore all the subdirectories of a directory _except one_. I put in .gitignore:18:02
*/18:02
!DirNotToIgnore/18:02
→ Diemuzi joined18:02
gitinfo set mode: +v18:02
pl it adds DirNotToIgnore, but not the files in it. Why?18:02
← netto_ left18:02
dAnjou hi, how can i view the changes of a file in a single commit? i have the commit sha and the filename.18:02
→ kpease joined18:02
Eugene dAnjou - `git diff SHA SHA^ -- filename`18:04
→ parcs joined18:04
parcs hi18:04
→ zackiv31 joined18:04
parcs why does git attempt to auto pack my repository after every operation?18:04
Eugene parcs - because your repository needs packing18:05
parcs why doens't a single auto pack suffice?18:05
zackiv31 why doesn't a git revert revert the git blame history as well? shouldn't it be able to tell me who the original creator of the lines were before someone messed up a commit ?18:05
→ hkon_ joined18:05
parcs even a nop "git fetch" triggers an auto pack..18:05
i read that if there are more than 27 files in .git/objects/17 then git will trigger a gc18:06
→ thibaultcha joined18:06
osse parcs: maybe you have accidentally configured the auto-packing somehome. Have you checked ~/.gitconfig and repo/.git/config ?18:06
parcs i have 38 objects in that directory18:06
and it remains at 38 even after each pack18:06
osse try a manual git gc --aggressive18:07
parcs ok... that's gonna take forever :(18:07
→ JosephFerano joined18:07
osse so be it18:08
hmm, maybe there are some permission problems that make git unable to remove objects it doesn't need anymore?18:08
→ AKK9 joined18:09
dAnjou Eugene: thanks. just tried something and this works too and even shows the meta data: git show d088c03 templates/client/page_project.html18:09
→ KevinTS joined18:09
dAnjou and it's more intuitive/straight forward18:09
→ crose joined18:09
← darkxploit left18:09
Eugene show will give you the contents, rather than the diff ;-)18:09
→ eletuchy joined18:09
parcs nope, everything is owned by me18:09
xender pl: Maybe you need !DirNotToIgnore/* ?18:10
pl: A for why - I dunno18:10
*as18:10
dAnjou Eugene: what's the difference in this case? i can see a diff here.18:10
← hkon_ left18:10
Eugene Oh, you used a space rather than a :18:10
pl xender, I tried, it's not that, but I think that the problem is that DirNotToIgnore is another git repo18:10
Eugene Nevermind, you've got the right one18:10
→ Tuna-Fish joined18:10
dAnjou :)18:10
xender pl: ...18:11
pl: !xy18:11
gitinfo pl: Woah, slow down for a bit. Are you sure that you need to jump through that particular hoop to achieve your goal? We suspect you don't, so why don't you back up a bit and tell us about the overall objective...18:11
parcs hmm maybe it's a bug in git 2.1. i am going to try git 2.018:11
→ listochkin joined18:11
osse parcs: gc --aggressive didn't reduce it?18:11
← hoschi left18:12
parcs i will try that next18:12
← Gizburdt left18:12
← crose left18:12
→ crose joined18:12
← eletuchy left18:13
← nnyk_ left18:14
→ Gizburdt joined18:14
osse dAnjou: you might want to write that git diff command the other way around18:15
← mathieuhays left18:16
→ mathieuhays joined18:17
→ ascarter_ joined18:17
→ pandeiro joined18:17
→ davispuh joined18:18
← ascarter left18:19
parcs well git gc with git 2.0 increased the number of loose objcets from 38 to 41...18:19
→ bgerber joined18:19
← AKK9 left18:19
parcs time to try --aggressive. see you in 30 minutes :(18:20
→ jottr joined18:20
→ kotk joined18:20
→ AKK9 joined18:21
← mathieuhays left18:21
→ zapb joined18:21
← kotk_ left18:22
← FortySix2 left18:22
← ntio left18:23
← Ankhers left18:23
← asakura1 left18:23
← skorgon left18:23
→ hebz0rl joined18:23
osse parcs: I seem to remember someone writing to the mailing list about the same issue you're seeing now18:23
but I can't find it18:23
← pnsrc left18:23
parcs good to know i'm not the only one..18:23
→ zumba_addict joined18:24
→ skorgon joined18:24
← codeFiend left18:24
→ asakura1 joined18:24
→ ntio joined18:24
→ Sleepy63 joined18:25
← _Vi left18:25
→ ninsei joined18:25
→ FortySix2 joined18:25
→ causasui joined18:26
← Relsak left18:29
→ silverf0x joined18:29
→ upene joined18:33
← upene left18:33
→ upene joined18:33
→ GodEater joined18:34
← GodEater left18:34
→ GodEater joined18:34
← jovy88 left18:34
→ KamiNuvini joined18:34
← jottr left18:35
→ FreezingCold joined18:35
← Xethron left18:35
→ jovy88 joined18:35
→ pnsrc joined18:36
← Mobutils left18:36
KamiNuvini Hi - I'm fairly new to git and I did a merge that was pushed to github that I want to revert. So with the merge things were placed in 2 different subdirectories. I wanted to split a subfolder into a new repository with the history - and once that is done reset the "main" repository back to a commit before the merge. I tried to follow the steps here:18:38
https://help.github.com/articles/splitting-a-subfolder-out-into-a-new-repository which seems to work but there aren't any instructions there on how to actually add that as a new repository. When I try to open that folder in Github (the windows program) it thinks it is still part of the old repository. How do I mark it as a "new" one?18:38
→ ferreum joined18:38
← nerdfiles left18:39
← lsiv568 left18:39
← listochkin left18:39
qsuscs step one: stop using github for windows18:39
bremner !gui18:40
gitinfo 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.18:40
→ nerdfiles joined18:40
→ lmat joined18:40
← thibaultcha left18:40
qsuscs besides, there _are_ good GUIs. github for windows not belonging to them.18:40
→ thibaultcha joined18:40
→ sinkensabe joined18:40
lmat when i git rebase --interactive --onto HEAD^^ HEAD^^ HEAD; my current branch doesn't move, only HEAD moves18:40
→ goTAN joined18:40
lmat OH YEAD!18:40
YEAH I need to do something like git --interactive --onto HEAD^^ HEAD^^ <branch>;18:41
KamiNuvini sure - the github for windows does have a shell as well. Does that work?18:41
→ Simgate joined18:41
qsuscs KamiNuvini: better, yes. you might want to change it to bash (default is windows powershell, which is pretty weird, especially for us here)18:42
KamiNuvini Ah wait. I think the easiest is if I just do it through linux. I moved stuff over to windows after that but I followed the steps initially via linux18:43
Okay so let's say I just followed these steps: https://help.github.com/articles/splitting-a-subfolder-out-into-a-new-repository - how do I save that as a new repository?18:44
→ cagedwisdom joined18:44
← goTAN left18:44
← ropav left18:45
← kermit left18:45
← sinkensabe left18:45
→ nemesit|znc joined18:45
← enki left18:45
→ LeMike joined18:46
→ Fissure joined18:46
→ msmto joined18:47
→ Xethron joined18:47
→ listochkin joined18:47
cbreak lmat: HEAD is your current branch18:48
lmat: by definition18:48
(current branch is defined as "what ever branch HEAD points at ")18:48
lmat cbreak: is one of your current branches by definition?18:48
cbreak no18:48
lmat cbreak: oh18:48
← ascarter_ left18:49
cbreak there's only one single current branch18:49
or none :)18:49
→ jerius joined18:49
lmat cbreak: I see... I forgot that HEAD doesn't just point to a hash, it points to refs.18:49
cbreak (detached head state of course)18:49
lmat nice18:50
cbreak: thanks.18:50
← FortySix2 left18:50
→ freezeve_ joined18:50
→ dreiss joined18:50
← freezevee left18:51
← kriks left18:51
→ ascarter joined18:53
← ascarter left18:53
→ ascarter joined18:54
← nerdfiles left18:54
→ jluc joined18:56
Nevik osse: hai18:56
→ ropav joined18:57
← ropav left18:57
→ FortySix2 joined19:00
→ plitzenberger joined19:00
← KevinTS left19:00
→ martinklepsch joined19:01
→ nisseni joined19:01
→ bezet joined19:01
← diegoviola left19:02
← Corey84 left19:02
← rolledback left19:02
→ _Vi joined19:02
← plitzenberger left19:04
← Xethron left19:05
← bezet left19:05
← Gizburdt left19:06
← lmat left19:06
→ Gizburdt joined19:06
← freezeve_ left19:07
← epifanio left19:07
→ LucasTT_ joined19:07
→ freezevee joined19:07
→ diegoviola joined19:08
← Diemuzi left19:08
→ lmat joined19:08
← pnsrc left19:08
← dccc left19:09
→ duderonomy joined19:09
← LucasTT left19:10
Simgate Hello ! I'm working on a project, and I would like to know if you guys would be interested by what I'm doing. It would be a service (like github), which would automatically build debian packages, rpm, ... of the git repo(s) of your choice each time someone commits something, and upload it somewhere that would be accessible by everyone. All you would have to do is tell that service what are the19:10
dependencies of your project, and what you want to put in the packages you want to create (among other things, but nothing inhuman). What would you say about such a thing ?19:10
← zapb left19:10
← Gizburdt left19:10
→ t4nk655 joined19:11
→ snuffeluffegus joined19:11
→ kriks joined19:11
← freezevee left19:11
→ epifanio joined19:12
← EminenceHC left19:12
→ EminenceHC joined19:12
gitinfo set mode: +v19:13
t4nk655 hello! i'm having issues connecting to github from my server (getting the "Permission denied (publickey). fatal: The remote end hung up unexpectedly" error), but it was working a couple days ago and I have not touched my public key on the server.. any ideas why this might be?19:13
→ nerdfile1 joined19:13
→ przemkovv joined19:13
← nisseni left19:13
cbreak t4nk655: try to ssh into github19:14
→ kermit joined19:15
t4nk655 thanks for the response! same issue (Permission denied (publickey).19:15
cbreak then your key isn't accepted19:15
try with more -v to see if the key is tried19:16
should give you something like debug1: Offering RSA public key: /Users/cbreak/.ssh/id_rsa19:16
gitinfo set mode: +v19:16
→ purpleidea joined19:16
verne.freenode.net set mode: +v19:16
t4nk655 yup, i'm getting that19:17
gitinfo set mode: +v19:17
cbreak so check if your account on github contains the matching public key19:17
← rgr left19:17
← EminenceHC left19:17
← dAnjou left19:17
t4nk655 k, thanks that's what i figured (don't currently have the account that manages the git repo..). just stumped as to what would have changed since i believe i'm the only one working on this right now.. thanks again!!19:18
← montchr left19:18
cbreak t4nk655: no account? That'd explain it :)19:18
t4nk655: you can only use ssh to access github if you have an account over there19:18
→ jfelchner joined19:18
t4nk655 sorry, to clarify, i have an account just don't have access to the one that manages the repo (so i can't check public keys)19:19
→ kballou joined19:20
→ jfelchne_ joined19:20
→ P4Titan joined19:21
→ beneggett joined19:22
→ Sonderblade joined19:22
P4Titan Hello all, I am an avid git user and am begninng to want to try my hand at some development for the git project. The git-scm.com website mentions a little about development and how it goes, coding styles and how to send a patch, etc... I am wondering how I could actually get involved, where to start, etc...19:23
← t4nk655 left19:23
← jfelchner left19:23
qsuscs P4Titan: see Documentation/SubmittingPatches in git.git19:23
→ r00tWearsRedHats joined19:24
→ joelteon joined19:24
→ Xethron joined19:24
← Xethron left19:24
→ Xethron joined19:24
joelteon is it a good idea to use something like git-encrypt, or should I just avoid checking that stuff into the source tree?19:24
P4Titan qsuscs: That only explains how to send a patch, what the procedure is, I am looking for a starting point19:24
→ dik_dak joined19:25
→ B1x1t0 joined19:25
qsuscs P4Titan: ah, i see. *shrug* as always, if you encounter a bug, fix it. if you miss functionality, propose/add it. cure your itch19:25
P4Titan Yes, but the thing is, I have not found a bug, nor do I see any missing functionality. But as joelteon pointed out, I could take a look at git-encrypt19:26
joelteon well, i wasn't suggesting it, i came in to ask19:26
i just arrived here19:26
→ sq-one joined19:27
P4Titan Oh, I thought you were recommending it for me, ha19:27
parcs osse: git gc --aggressive didn't help19:29
← wild_oscar left19:29
parcs what an immense waste of electricity19:29
← esclear left19:30
→ mspe joined19:30
→ rolledback joined19:31
→ thiago_ joined19:31
← dogeydogey left19:32
→ zapb joined19:32
← skorgon left19:33
→ skorgon joined19:33
← thiago left19:33
→ DeathIncarnate joined19:36
← Xethron left19:37
ojacobson joelteon: git-encrypt only works for _anything_ if you have 100% perfect key hygiene, since there's no way to re-key old commits without invalidating their hashes.19:39
Since you don't have perfect key hygiene, it can't possibly work for anything.19:39
joelteon yeah, that's what I was afraid of19:39
ojacobson What were you thinking of encrypting?19:39
joelteon eh, I have some puppet manifests with some SSH keys inside19:39
← robarelli left19:39
bremner public or private keys?19:40
→ nnyk_ joined19:40
joelteon private19:40
bremner yuck19:40
joelteon oh wait, not ssh keys19:40
ssl certs19:40
duh, i knew something sounded wrong19:41
→ jatt joined19:41
← jatt left19:41
→ jatt joined19:41
→ MattMaker joined19:43
← Quevin left19:43
ojacobson joelteon: we split our puppet data into two repos19:44
→ robarelli joined19:44
ojacobson one has data-only classes, the other has normal classes with no secrets (that accept hiera parameters designating which secret to use)19:44
joelteon yeah, see, our work puppet masters don't have hiera19:44
ojacobson :-|19:45
If you're pre-2.7, fuck off and upgrade :)19:45
→ SJS joined19:45
joelteon ok, i'll forward that to them19:45
ojacobson if you're on 2.7, install puppet-hiera and follow the puppetlabs style guide, which will create manifests compatible with three19:45
and which will get you most of puppet 3's (Rather Nice) data binding bits19:45
→ gregsmith joined19:45
joelteon what would you do if we were using, say, ansible19:45
ojacobson without hiera though you could also hard-code the names of secrets (as $variable::names) and just assume the secrets tree were present19:46
→ Quevin joined19:47
← msmto left19:47
← martinklepsch left19:47
← listochkin left19:48
→ martinklepsch joined19:49
← jatt left19:52
← nerdfile1 left19:52
→ listochkin joined19:53
← przemkovv left19:53
→ axl_ joined19:55
← lpax left19:57
→ nerdfile1 joined19:58
← foist left19:59
→ steven_ joined19:59
→ adhawkins joined20:00
← steven_ left20:00
← kpease left20:00
← m0ltar left20:01
→ Sasazuka joined20:01
→ Tuxguy joined20:01
← beneggett left20:02
Tuxguy Is it possible to simulate a merge to see what the outcome would look like?20:02
→ lsiv568 joined20:02
← Zamerick left20:02
ToxicFrog Tuxguy: git merge --no-commit?20:02
← davispuh left20:02
→ phanimahesh joined20:02
gitinfo set mode: +v20:02
→ sniptras joined20:05
Tuxguy Toxic frog, I think I have to manually merge... I see some things in diff that i need in file_a_branch_a and file_a_branch_b and its not always one direction20:05
← treehug88 left20:05
← Technodrome left20:06
→ hkon_ joined20:06
→ binaryhat joined20:07
Tuxguy ToxicFrog: how can I do that? Like, force a manual merge of diffs?20:08
ToxicFrog What do you mean by "manual merge of diffs"?20:08
_ikke_ basically a two way merge20:08
→ EminenceHC joined20:09
ToxicFrog If you want to merge one branch into another, use git merge. If you want to merge both branches into each other, so that both branch heads point to the same merge commit, merge a b and then merge b a; the second time will fast-forward.20:09
_ikke_ ToxicFrog: He wants to manually select which hunk to choose from branch a or from b20:09
← crose left20:09
ToxicFrog If you want to "merge" specific changes from one branch into another without actually creating a merge commit, you actually want the cherry-pick command.20:09
Oooh.20:10
So basically force conflict resolution even if there are no conflicts?20:10
← rolledback left20:10
_ikke_ aye20:10
ie. two way merge20:10
→ plitzenberger joined20:10
← hkon_ left20:11
→ Drakevr joined20:11
_ikke_ You can try this: git merge -n <branch>; git checkout -p --theirs; git checkout -p ours20:11
Tuxguy: ^20:11
→ edude03 joined20:11
← edude03 left20:12
→ edude03 joined20:12
← jovy88 left20:12
→ kpease joined20:13
→ BSaboia joined20:13
→ KevinTS joined20:13
→ jovy88 joined20:13
← gyyrog left20:14
← FreezingCold left20:15
← asakura1 left20:16
← pheaver left20:16
← ninsei left20:16
← causasui left20:16
→ gattuso joined20:17
→ ninsei joined20:17
→ Mobutils joined20:17
→ causasui joined20:17
→ mathieuhays joined20:17
← KevinTS left20:17
← akushner left20:17
→ asakura1 joined20:17
→ codeFiend joined20:19
← codeFiend left20:20
→ Panicky joined20:21
← Quevin left20:21
→ sinkensabe joined20:22
← mathieuhays left20:22
→ codeFiend joined20:23
thiago_ → thiago20:24
← vurma left20:24
← sinkensabe left20:25
→ crose joined20:25
→ Technodrome joined20:25
→ yosito joined20:26
→ foist joined20:26
→ awkwardusername joined20:26
→ skfax joined20:27
Tuxguy _ikke_: ty, ill try that when i get home. im gonna email myself that :)20:27
→ netcarver joined20:27
skfax If I have a git repo locally, is it possible to push this to another new repo which is bare?20:28
tobiasvl yes of course20:28
→ veebers joined20:29
→ koki_ joined20:29
← iarna left20:29
Tuxguy _ikke_: worse case i checkout files in some tmp space from both branches that are in the git..diff and use a tool like kdiff/winmerge etc?20:29
→ iarna joined20:30
→ FreezingCold joined20:30
← foist left20:31
← listochkin left20:31
→ sinkensabe joined20:31
skfax tobiasvl: Found it now :) Easy solution was to just use git clone --bare ~/original from another location20:32
→ Kartagis joined20:32
Kartagis hi20:32
← JDHawke left20:33
→ JDHawke joined20:33
← skfax left20:33
← morphix0 left20:33
Kartagis how can I prepare a patch like the one at http://html.webciniz.im/106713.patch ?20:33
I mean, with all the info20:34
← Tuxguy left20:34
_ikke_ Kartagis: use git format-patch20:35
← sinkensabe left20:35
→ Wiz_KeeD joined20:35
← Simgate left20:35
← awkwardusername left20:35
→ looper joined20:35
← FortySix2 left20:36
Kartagis _ikke_: someone here (I don't remember who) told me it took two arguments; both commit hashes. since a commit hash can be generated only by a commit, can I prepare such a patch for someone else's project?20:37
ojacobson Sure. Create a commit.20:38
← LBV_User left20:38
← Mobutils left20:39
← nerdfile1 left20:40
→ nerdfile1 joined20:40
← sq-one left20:41
← P4Titan left20:43
→ netto_ joined20:44
← netto_ left20:45
→ LBV_User joined20:45
→ morphix0 joined20:46
← Wiz_KeeD left20:46
← nnyk_ left20:47
← pl left20:47
← sniptras left20:48
← looper left20:48
→ sniptras joined20:48
← DeathIncarnate left20:49
→ DeathIncarnate joined20:49
→ Pookz joined20:49
→ hj2014_ joined20:51
← hj2014 left20:51
← fr33domlover left20:51
← plitzenberger left20:52
→ staguer joined20:52
← Kicer86 left20:52
→ KevinTS joined20:53
→ FortySix2 joined20:56
← duderonomy left20:56
← morphix0 left20:57
→ dserodio joined20:57
→ bitkiller joined20:57
← asakura1 left20:58
← HanClinto_ left20:58
→ asakura1 joined21:00
← Xiti` left21:00
→ Xiti joined21:01
→ sinkensabe joined21:02
→ endiruna joined21:02
← zumba_addict left21:03
→ JohnBat26 joined21:03
→ zumba_addict joined21:03
→ alvaro_o joined21:03
← stamina left21:04
← KevinTS left21:06
→ listochkin joined21:06
→ KevinTS joined21:07
→ Mobutils joined21:07
← toedwy left21:08
← zumba_addict left21:08
← kermit left21:08
→ Sceorem joined21:08
→ tpollard joined21:09
← Sceorem left21:09
→ Quevin joined21:10
← ojacobson left21:11
← KevinTS left21:11
→ glebd joined21:11
→ kyb3r_ joined21:12
← jerius left21:12
← zapb left21:12
← qwyeth_ left21:13
← sdrfnord1 left21:13
→ nnyk_ joined21:13
← diegoviola left21:14
← nerdfile1 left21:14
→ akushner joined21:14
→ ItSANgo joined21:15
→ jottr joined21:15
→ Megaf_ joined21:15
← FortySix2 left21:15
← Megaf left21:16
← glebd left21:16
→ glebd joined21:16
← alexbligh1 left21:18
← glebd left21:18
← mspe left21:18
← kriks left21:18
→ alexbligh1 joined21:18
→ kermit joined21:18
← akushner left21:18
→ glebd joined21:18
ChanServ set mode: +o21:19
Eugene changed the topic to: Welcome to #git, the place for git help and flaky telephone circuits | Current stable version: 2.0.3 | Start here: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Now with more indirect cycling giraffes!21:19
← ItSANgo left21:19
Eugene set mode: -o21:19
xender \o/ I've found a way to spawn gitk with "ignore certain refs" option while not specifying desired refs at startup (which means updating list of refs without restarting. when branches are made or deleted)21:19
← lsiv568 left21:19
→ lsiv568 joined21:20
xender gitk --argscmd="git for-each-ref --format='%(refname)' | grep -v '<unwanted refs>'" HEAD21:20
argscmd is re-evaluated every time graph is to be refreshed21:20
→ the8thbit|work joined21:21
→ Vile joined21:21
→ ItSANgo joined21:22
← the8thbit|work left21:22
← BSaboia left21:22
parcs osse: the trick was to use "git gc --prune=all". git gc only prunes objects that are older than 2 weeks21:22
→ tvw joined21:22
← skylite left21:22
→ coldkey21 joined21:23
← lsiv568 left21:24
→ plitzenberger joined21:24
← Mobutils left21:24
← jottr left21:24
→ FortySix2 joined21:24
← FortySix2 left21:25
→ FortySix2 joined21:25
← epifanio left21:25
← joelrobichaud left21:26
← lmat left21:28
→ aartist joined21:28
→ guzzlefry joined21:28
← kevinb left21:29
→ epifanio joined21:29
→ sdrfnord1 joined21:29
← denisMone left21:30
→ ropav joined21:30
← upene left21:30
→ kevinb joined21:31
→ Reskp joined21:32
← carandraug left21:32
← ropav left21:33
→ skylite joined21:33
← skylite left21:33
← acharlieh left21:33
← sdrfnord1 left21:33
← DeathIncarnate left21:34
← mennea left21:34
→ upene joined21:34
← yosito left21:34
→ hj2014 joined21:35
→ MasH4077TH joined21:36
← JDHawke left21:36
→ salamanderrake joined21:37
← sniptras left21:37
← epifanio left21:38
→ acharlieh joined21:38
← hj2014_ left21:38
→ freezevee joined21:39
← LBV_User left21:40
→ lakcaj joined21:42
→ ropav joined21:42
→ jottr joined21:43
osse parcs: HMMMMMMMMMMMMMMMMMMMMMMMMM21:43
← ropav left21:43
osse parcs: that makes sense, but it a "how the hell should I have known that?" sort of way21:43
parcs i think the auto gc logic is pretty broken21:44
osse parcs: how did you find out?21:45
→ toedwy joined21:45
→ mablae_ joined21:45
mablae_ hi21:45
gitinfo mablae_: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.21:45
← Asiajey left21:45
→ rgr joined21:45
parcs i noticed that the git gc manual says that it only prunes loose objects that are older than 2 weeks21:45
then i checked the date on the loose objects (17 july)21:46
→ hj2014_ joined21:46
mablae_ I am using gitlab on LAN. Is it possible to push from gitlab to an external repo on changes?!21:46
← upene left21:46
parcs i wonder why they were dated 17 july though21:46
→ Mobutils joined21:46
→ Asiajey joined21:47
osse parcs: maybe you did a rebase or commit --amend that day21:47
any rewriting leaves a lot of garbage in its wake21:47
mablae_ My problem is the fact that gitlab only holds a bare repo.. How can push from that bare repo to an external repo?!21:47
parcs maybe.. i don't remember21:47
cbreak mablae_: git push21:47
parcs but suspiciously 17 july is 13 days ago21:47
cbreak mablae_: pushing doesn't need a working dir21:48
→ nerdfile1 joined21:48
← hj2014 left21:48
osse parcs: ahh, you think it's some sort of conspiracy?21:48
mablae_ So in the hook I just say "git push ..."21:48
parcs osse: and/or an off-by-one error21:48
mablae_ Is there some env vars set?!21:48
I didn't tried out yet...21:48
Maybe I should just try21:49
→ bob14 joined21:49
mablae_ cbreak: So I need just a post-update hook? Or post-receive? Where is the diff between them?21:49
bob14 → joe2321:50
← staguer left21:50
cbreak mablae_: post receive is after everything's done, all refs are updated21:50
probably the best spot21:50
← bitkiller left21:50
← jfelchne_ left21:50
mablae_ I need to push only the release branch to the upstream21:51
→ bitkiller joined21:51
← salamanderrake left21:51
→ BSaboia joined21:52
← jovy88 left21:53
→ hkon_ joined21:54
← Sleepy63 left21:55
→ jovy88 joined21:55
→ DeathIncarnate joined21:55
← daedeloth left21:57
← Mobutils left21:57
→ Bruno_ joined21:57
← sinkensabe left21:58
← Arkaniad left21:58
← nerdfile1 left21:58
← jottr left21:59
← hkon_ left21:59
← Dougie187 left22:00
→ bambams joined22:01
← macabre left22:01
← fstd left22:01
→ fstd joined22:02
→ ropav joined22:02
← a_le left22:02
← Gege left22:03
→ nafg_ joined22:03
→ a_le joined22:03
← ropav left22:03
cbreak so?22:04
you can push what ever you want.22:04
→ salamanderrake joined22:04
cbreak you might want to check if release was updated in the hook though, if you want to avoid unneeded pushes22:04
← Reskp left22:04
→ Reskp joined22:05
→ Gege joined22:06
← a_le left22:07
→ Sleepy63 joined22:07
→ jason237 joined22:07
← sonicparke|bradm left22:08
← Reskp left22:10
← crose left22:10
mablae_ cbreak: Yeah I found that22:10
→ Reskp joined22:11
mablae_ But: When I am in the hook, how would I say that only the release branch should be pushed? With no working dir, there is no current branch checked out, right..22:12
← causasui left22:12
→ ropav joined22:12
→ causasui joined22:12
cbreak mablae_: so?22:13
← sohail_ left22:13
cbreak git push upstream master22:13
just like normal22:13
← Gege left22:13
→ cmosguy joined22:14
← cmosguy left22:14
← joe23 left22:14
→ cmosguy joined22:14
→ akushner joined22:15
gitinfo set mode: +v22:15
cmosguy i accidentally destroyed a sub-directory under my project, how do I restore it from the repo?22:15
→ freestyl3r joined22:15
mablae_ but normally in non-bare there is master checked out locally...22:15
← fatalhalt left22:15
← ropav left22:16
osse cmosguy: git checkout HEAD -- sub-directory22:16
mablae_ cbreak: Or is the "master" for source and dest in this case?!22:16
cbreak mablae_: master is both source and destination22:16
the assumption is that you want to call the branch the same on both ends22:16
← FortySix2 left22:17
cbreak mablae_: if you want to call it differently, use git push remotename foo:bar22:17
→ ropav joined22:17
cmosguy mablae_, woosh that was close! thanks for the help!22:17
← martinklepsch left22:18
← MasH4077TH left22:18
mablae_ cmosguy: it was osse helping you22:18
I am dum noob, in git :)22:18
→ mathieuhays joined22:18
cmosguy osse, thanks!22:18
mablae_ cbreak: Yeah, I just tried... Works great! And if the branch is not yet in dest? Add -u param?!22:19
cbreak no22:19
← akushner left22:20
cbreak it'll be created22:20
← veebers left22:22
→ knighthawk joined22:23
← mathieuhays left22:23
mablae_ Ah, okay!22:23
Thanks for your help...22:23
→ Arkaniad joined22:24
mablae_ Do you use gitlab? Is it save to give the gitlab user an own id_rsa for auth againt external repos?22:24
← nafg_ left22:25
← abra left22:26
→ Sharcho joined22:28
← Sonderblade left22:28
→ abra joined22:28
→ cfxd joined22:29
← acharlieh left22:30
← tranzfus_ left22:30
← ropav left22:30
→ tranzfuse joined22:31
← JohnBat26 left22:31
← pantsman_ left22:31
→ mribeirodantas joined22:31
→ mania joined22:31
← BSaboia left22:31
← Vile left22:32
← Destos left22:33
→ jjasinski joined22:33
← B1x1t0 left22:34
← nnyk_ left22:34
← AKK9 left22:35
→ ropav joined22:35
← tranzfuse left22:35
← Neozonz|Disc left22:36
← mania left22:37
← ropav left22:38
→ BSaboia joined22:39
← ferreum left22:39
→ lidaaa joined22:40
→ fold joined22:40
← delight left22:42
SamB wonders if <http://permalink.gmane.org/gmane.comp.version-control.git/254055> qualifies as spam22:43
← hebz0rl left22:43
← nickpresta left22:43
→ delight joined22:44
→ jottr joined22:44
→ reaverb joined22:45
SamB so, um, cache.h and git-compat-util.h are IMO fairly annoying because they each seem like they should actually be a bunch of smaller headers ...22:45
← it0a left22:46
→ Gizburdt joined22:46
← Gizburdt left22:46
→ msmto joined22:46
osse hates cache.h22:47
osse and sometimes when you include another header because you need somethnig from it you get unrelated compilation errors22:47
← plitzenberger left22:47
← Tuna-Fish left22:47
osse and guess what: including cache.h before it fixes the problem. grrrr22:47
← dreiss left22:47
qsuscs oh, new version, 2.0.422:48
This is primarily to fix a regression of "git diff-tree" in v2.0.2.22:48
→ ropav joined22:50
← Sasazuka left22:51
← thedodd left22:51
→ veebers joined22:51
← tummy left22:51
← abra left22:53
← ropav left22:54
← phanimahesh left22:54
→ abra joined22:55
← DeathIncarnate left22:56
→ dsdeiz joined22:57
→ Tuna-Fish joined22:58
→ fstd_ joined22:58
← cmosguy left22:59
→ Textmode joined22:59
← jluc left22:59
← codeFiend left22:59
→ misterpink joined23:01
← delight left23:01
← hcc23 left23:01
→ tranzfuse joined23:01
→ bezet joined23:02
← sgambino left23:02
← fstd left23:02
fstd_ → fstd23:02
→ Gizburdt joined23:03
→ a_le joined23:03
→ ropav joined23:03
Vampire0 .version23:03
gitinfo changed the topic to: Welcome to #git, the place for git help and flaky telephone circuits | Current stable version: 2.0.4 | Start here: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | Now with more indirect cycling giraffes!23:03
→ thedodd joined23:05
→ p014k joined23:05
← Vivekananda_y510 left23:06
← tranzfuse left23:06
← ropav left23:06
← bezet left23:06
→ Vivekananda_y510 joined23:06
← punchagan left23:07
→ przemkovv joined23:08
p014k I'd like to clean up my commit history a little bit be completely deleting several commits. I can't seem to google how to. I've seen 'git rebase -p --onto SHA^ SHA', but I'm unclear if this is correct or if it will work.23:08
← rgr left23:08
osse !sausage23:08
gitinfo [!sausage_making] Some developers like to "hide the sausage making", transforming their commits before presenting them to the outside world. See http://sethrobertson.github.com/GitBestPractices/#sausage and !perfect23:08
← tvw left23:08
osse p014k: your rebase will remove SHA from the history, but that's about it23:09
→ andkore joined23:09
osse p014k: you probably want git rebase -i HEAD~n where n is a sufficently high number to cover all the commits you want to fiddle with23:10
p014k Ya, that's all I want to do. Remove it from history. The current HEAD is good, but I'd just like to delete some pointless commits it took to get there (here).23:11
andkore Is Git the best option for storing the history of a document? Is there a good way to get around the one repo per directory design?23:11
← tang^ left23:11
osse andkore: it is a good option, but I hesitate to say the best. No.23:12
p014k going back to the rebase command, what is the -p --onto SHA^ SHA option actually doing?23:12
andkore osse: It seems pretty inconvenient to make a new directory for every document. What would you recommend instead?23:12
→ ropav joined23:12
osse andkore: I don't have any other good recommendation, to be honest.23:12
← cfxd left23:12
osse most of my documents (to the extent I write any) tend to consist of several files23:13
so I haven't given it much thought23:13
→ Balliad joined23:13
osse p014k: git rebase SHA rebases all commits reachable from HEAD but not reachable from SHA onto SHA23:14
p014k: when you add --onto SHA^ the same commits are rebased, but onto SHA^ instead, effectively leaving SHA itself out of the picture23:14
andkore osse: OK, thanks.23:14
← ropav left23:15
p014k Thanks!23:15
osse andkore: it is possible with some shell fiddling to create a wrapper that uses .git for one repo's content (as usual) but uses .git2 or .git-other-stuff for another23:15
I mean: It's technically possible, but I don't know of any existing solution23:15
→ akushner joined23:15
andkore Yeah. I think I'll just make new directories. I guess it kind of makes sense anyway.23:16
osse andkore: i'm curious since it's always only one file: what kind of document ?23:16
andkore osse: Right now it's a .tex (LaTex) file. I don't use Word much these days, but when I was in school it would have been nice to have version histories for my papers.23:17
osse andkore: no figures then I guess :)23:18
→ freezeve_ joined23:18
andkore osse: Philosophy.23:18
← freezevee left23:18
osse ahh, so lots and lots and lots of words then23:18
maybe you could consider splitting it up in several tex files :P23:18
\include{introduction.txt} and \include{conclusion.tex} or what-have-you23:19
← kevinb left23:19
andkore Right. This is just a resume so it doesn't really need to be split up.23:19
← Gizburdt left23:19
→ akushner_ joined23:19
osse I remember something about a convention being that you had a file called 'master.tex' that included everything else you needed.23:19
→ jfelchner joined23:19
← akushner left23:20
andkore It's kind of nice to have all of your documents in one directory/folder so you can see them all together. But I guess having the history is probably worth losing that.23:20
→ ropav joined23:20
osse I miss latex23:21
andkore Anyway, thanks for the help osse. I kind of figured that there wasn't a better solution, but I just wanted to make sure.23:21
I just learned a little about it.23:21
osse i used it a lot at uni23:21
andkore The resume I'm working on looks way better than the resumes I've made in the past.23:21
← msmto left23:21
← joelm left23:21
osse some a resume qualifies as philosophy these days? :P23:22
← knighthawk left23:23
andkore No, I said I wrote my papers in Word, and it would have been nice to have the file histories.23:23
But that right now it's a LaTex file.23:23
← ciampix__ left23:23
→ LucasTT2 joined23:23
→ HanClinto_ joined23:23
← axl_ left23:24
osse ooh23:25
Word itself has a kind of history/review feature built-in23:25
← ropav left23:25
osse It always confuses me though23:25
andkore Yeah, right.23:25
I prefer solutions that aren't that specific.23:26
→ kevinb joined23:26
← nfk left23:27
← LucasTT_ left23:27
andkore And just so you know, one of the most influential philosophy articles in the 20th century was about two or three pages.23:27
Philosophy isn't about writing pages and pages of nonsense.23:27
→ foist joined23:27
thibaultcha What is this article andkore?23:27
andkore thibaultcha: Gettier's paper. Lemme look up the title.23:28
→ shinnya joined23:28
→ Mobutils joined23:28
andkore "Is Justified True Belief Knowledge?" That's what I thought the title was...23:28
http://philosophyfaculty.ucsd.edu/faculty/rarneson/Courses/gettierphilreading.pdf23:29
thibaultcha Is it this: http://fitelson.org/proseminar/gettier.pdf ?23:29
andkore It's l ike 2.5 pages.23:29
Yes, that's it.23:29
like*23:29
thibaultcha Ok thank you, I was curious. I'll read it.23:29
→ flaguy48 joined23:30
osse andkore: heh I was just winding you up23:30
← jottr left23:30
andkore osse: Yeah. I'm kind of prone to being wound up about philosophy because most people have no idea what it's actually about and think it's just bullshit.23:30
osse but I remember reading a paper about determinism. And remember each paragraph was much longer what I was used to, and that many sentences were so long that I had forgotten how they started by the time I got to the end23:31
granted, this was in high school23:32
← Embassy left23:32
→ ropav joined23:33
osse it would be fun to try to read it again. it was way over my head at the time.23:34
andkore osse: Yeah, I'm kind of glad I didn't have any philosophy classes in high school.23:34
← robarelli left23:35
→ Apocalisp joined23:35
osse oh I just found it online. It wasn't in my curriculum or anything23:35
andkore Ah.23:35
← gregsmith left23:36
→ tmcmahon joined23:36
→ JakeSays joined23:37
→ nickpresta joined23:38
← a_le left23:38
→ a_le joined23:39
← Skapadapaduu left23:39
← asakura1 left23:40
→ maslowbeer joined23:42
← ropav left23:42
→ asakura1 joined23:42
← mikecmpbll left23:42
← knownasilya left23:43
→ hkon_ joined23:43
maslowbeer i've read that git-subtree has been merged into mainline git but I don't have that command (v. 1.8.1.2). What am I missing?23:43
← smw left23:44
→ Gege joined23:44
maslowbeer gitinfo, help23:44
osse !subtree23:44
gitinfo The subtree merge method is great for incorporating a subsidiary git repo into your current one with "unified" history. Read http://git-scm.com/book/en/Git-Tools-Subtree-Merging for more info, see also !git-subtree and !git-stich-repo.23:44
osse !git-subtree23:44
gitinfo git-subtree allows a subproject repository to be incorporated into a subdirectory of a main repository, and for the history of a subdirectory to be (re-)exported with reproducible results, making it possible to synchronise a repository with a subdirectory of another repo, see https://github.com/git/git/blob/master/contrib/subtree/git-subtree.txt for more info.23:44
osse that last one there23:44
SamB what's with "unsigned long" for inode numbers?23:44
← skorgon left23:45
← dik_dak left23:45
→ skorgon joined23:45
← tmcmahon left23:45
← FreezingCold left23:46
→ godgodgodgo joined23:46
← Balliad left23:47
→ ropav joined23:47
← hkon_ left23:47
← thedodd left23:48
→ Lyqyd joined23:49
→ plitzenberger joined23:50
→ kukks joined23:50
← kballou left23:51
SamB hmm, how do you printf an off_t ?23:51
I guess it'd have to be a macro, wouldn't it?23:51
→ thedodd joined23:52
→ Destos joined23:53
← Mobutils left23:53
← godgodgodgo left23:54
← listochkin left23:56
← Reskp left23:56
→ Reskp joined23:57
← thedodd left23:57
maslowbeer so basically what I'm trying to do is this: I have a repo already on my local. I have a remote foo in addition to origin. I want to just push a single folder in my repo to the foo remote. I was thinking either a recursive submodule or subtree if that would do the trick. thoughts?23:58
← nickpresta left23:59
→ sinkensabe joined23:59
→ dogeydogey joined23:59

Logs Search ←Prev date Next date→ Channels Documentation