Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-10-06

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

All times shown according to UTC.

Time Nick Message
00:02 red82 joined #git
00:03 jaguarmagenta joined #git
00:04 harish joined #git
00:06 iveqy joined #git
00:06 iveqy_ joined #git
00:08 eka joined #git
00:10 aswen joined #git
00:11 aeae64ff8df64d2a joined #git
00:11 aswen joined #git
00:13 Anja joined #git
00:14 fuchstronaut joined #git
00:16 mehola joined #git
00:17 TinkerTyper joined #git
00:17 mattcen joined #git
00:17 qpdb joined #git
00:25 bisch joined #git
00:26 jason237 joined #git
00:27 iveqy joined #git
00:27 iveqy_ joined #git
00:28 underpantz__ joined #git
00:30 mattcen joined #git
00:32 perrier-jouet joined #git
00:33 Anja joined #git
00:33 mizu_no_oto joined #git
00:33 iveqy joined #git
00:34 iveqy_ joined #git
00:36 NeverDie joined #git
00:36 kerrick joined #git
00:38 cspack_ joined #git
00:38 dendazen joined #git
00:39 prawnzy joined #git
00:39 nidr0x joined #git
00:39 raijin joined #git
00:43 navidr joined #git
00:47 Anja joined #git
00:50 ahr3n joined #git
00:50 mattcen joined #git
00:51 EmberCrest joined #git
00:51 EmberCrest Ok I'm about to initially commit my project to Github for the 5th time this week because my MySQL password keeps ending up in the repo.
00:52 EmberCrest I feel like a dumbass, but I keep trying various solutions to this problem to their eventual failure.
00:52 EmberCrest I need to commit the file "Server.Settings" _ONCE_ and then I need git to forever ignore it..
00:52 EmberCrest can anyone save me from my self-inflicted spiral of retardation?
00:53 crayon joined #git
00:53 crayon joined #git
00:53 rominronin joined #git
00:54 Vampire0 EmberCrest, sure: don't do it
00:55 Vampire0 EmberCrest, !config
00:55 gitinfo EmberCrest: [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas
00:55 EmberCrest Ok that's not really an option for me, but thanks for the tip
00:56 dsantiago joined #git
00:56 star_prone joined #git
00:58 EmberCrest git update-index --assume-unchanged
00:58 EmberCrest This is what I needed, never mind thanks :)
00:58 EmberCrest left #git
00:59 bremner classic git: here is this shiny thing that will probably hurt you. Have fun!
00:59 LeBlaaanc joined #git
00:59 ahmed_elgabri joined #git
01:00 Darren_ joined #git
01:01 e14 joined #git
01:01 Anja joined #git
01:02 justan0theruser joined #git
01:03 meloc joined #git
01:04 Vampire0_ joined #git
01:07 star_prone joined #git
01:07 harish joined #git
01:10 nd joined #git
01:11 nettoweb joined #git
01:15 fuchstronaut joined #git
01:16 ToBeCloud joined #git
01:23 raijin joined #git
01:23 sunri5e joined #git
01:26 d^sh joined #git
01:26 fstd_ joined #git
01:29 cdg joined #git
01:31 pks joined #git
01:32 fstd joined #git
01:33 cyan__ joined #git
01:34 raijin joined #git
01:34 rpg joined #git
01:37 Kitty- joined #git
01:37 star_prone joined #git
01:38 rtm1046 joined #git
01:39 mattcen joined #git
01:46 Dazzyp joined #git
01:46 cagedwisdom joined #git
01:46 LBV_User joined #git
01:48 ilbot3 joined #git
01:48 Topic for #git is now Welcome to #git, the place for git help and electronically-induced comas | Please take the git survey: https://jk.gs/gsurv | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.10.1 | Getting "cannot send to channel"? /msg gitinfo .voice | git-daemon: the evil rises
01:50 cyphase joined #git
01:53 fstd joined #git
01:57 duderonomy joined #git
01:58 Lipp joined #git
02:01 Limes_ joined #git
02:01 mehola joined #git
02:02 mizu_no_oto joined #git
02:02 cyphase joined #git
02:04 jaguarmagenta joined #git
02:06 Lyqyd joined #git
02:08 |ifei5g00d joined #git
02:11 boombatower joined #git
02:12 |ifei5good joined #git
02:14 raspado joined #git
02:14 rtm1046 joined #git
02:16 jaguarmagenta joined #git
02:19 anonymuse joined #git
02:19 dreiss joined #git
02:20 kyledr joined #git
02:21 ochorocho__ joined #git
02:21 rtm1046 joined #git
02:22 kerrick joined #git
02:25 dsantiago joined #git
02:25 rominronin joined #git
02:27 johnmilton joined #git
02:30 red82 joined #git
02:32 mattcen joined #git
02:34 jimi_ joined #git
02:34 SpeakerToMeat joined #git
02:36 CEnnis91 joined #git
02:37 tehtraveler2321 joined #git
02:41 kerrick joined #git
02:41 ahr3n joined #git
02:44 red82 joined #git
02:45 iveqy joined #git
02:45 iveqy_ joined #git
02:46 hexagoxel joined #git
02:47 alex1a joined #git
02:48 sunrunner20 joined #git
02:48 sunrunner20 warning: venting
02:48 sunrunner20 what the hell is so hard about *basic* Git usage
02:49 sunrunner20 seems like half the developers at work can't grasp: Pull, merge, commmit, push
02:49 sunrunner20 they don't even have to deal with branches!
02:50 nedbat sunrunner20: what vcs are they used to?
02:50 sunrunner20 minus the initial "checkout development"
02:50 sunrunner20 nedbat: one of them VSS, the other "a varity, but mostly SVN"
02:51 sunrunner20 I used SVN for years, then stoped all tech for like 5
02:51 sunrunner20 came back
02:51 sunrunner20 *5 years
02:51 sunrunner20 git was the RCS of choice
02:51 nedbat sunrunner20: i'm not sure I count VSS as a version control system :)
02:51 sunrunner20 ha
02:51 sunrunner20 agreed
02:52 sunrunner20 still used by one group at work
02:52 sunrunner20 because they can't lose the history on the files
02:52 nedbat yup
02:52 sunrunner20 and they're truely indipendent files not a project or something
02:52 sunrunner20 I think I found something that can import the history
02:52 sunrunner20 I'll try it out eventually and see what they thing
02:53 sunrunner20 *think
02:53 sunrunner20 anyway
02:53 sunrunner20 I came from SVN
02:53 hahuang62 joined #git
02:53 sunrunner20 took me 3hrs and somebody showing it to me a couple times to be able to consistantly check stuff in and push it
02:53 sunrunner20 *of
02:54 nedbat sunrunner20: git definitely has a learning curve
02:54 fedepad joined #git
02:59 chachasmooth joined #git
03:02 Anja joined #git
03:05 mehola joined #git
03:09 kbdkode joined #git
03:10 pikajude_ joined #git
03:11 star_prone joined #git
03:11 kbdkode joined #git
03:12 pikajude_ joined #git
03:12 armyriad joined #git
03:13 Anja_ joined #git
03:14 Vortex35 joined #git
03:15 watersoul joined #git
03:16 fuchstronaut joined #git
03:17 cdg_ joined #git
03:18 rgrinberg joined #git
03:19 iveqy joined #git
03:20 iveqy_ joined #git
03:20 whitby joined #git
03:20 Lipp joined #git
03:22 ochorocho__ joined #git
03:26 snthsnth joined #git
03:27 duderonomy joined #git
03:28 Kitty- joined #git
03:29 nettoweb joined #git
03:29 steven_a_s joined #git
03:30 ahmed_elgabri joined #git
03:31 nckpz joined #git
03:33 einnjo joined #git
03:35 cagedwisdom joined #git
03:36 nettoweb joined #git
03:36 kerrick joined #git
03:37 BenderRodriguez joined #git
03:39 Darren_ joined #git
03:40 raspado joined #git
03:41 steven_a_s joined #git
03:44 Opertus joined #git
03:46 Opertus joined #git
03:46 kerrick joined #git
03:51 greendo joined #git
03:53 safe joined #git
03:54 Andrew_K joined #git
03:56 SwiftMatt joined #git
03:57 rominronin joined #git
03:59 mizu_no_oto joined #git
04:00 ZeroTresBR joined #git
04:03 Mobutils_ joined #git
04:03 daey_ joined #git
04:05 cagmz joined #git
04:05 nettoweb joined #git
04:12 raspado joined #git
04:16 Duder9999 joined #git
04:16 phanimahesh joined #git
04:17 jaguarmagenta joined #git
04:18 bmahe joined #git
04:19 eizua joined #git
04:19 Duder9999 joined #git
04:24 dreiss joined #git
04:24 einnjo joined #git
04:26 eshlox joined #git
04:26 Jameser joined #git
04:28 mattcen joined #git
04:29 jstimm joined #git
04:33 Lipp joined #git
04:41 jaequery joined #git
04:42 atomi joined #git
04:46 nevodka joined #git
04:48 eizua joined #git
04:49 mangolisk joined #git
04:51 auzty joined #git
04:52 Andrew_K joined #git
04:54 Cabanossi joined #git
04:55 nilg joined #git
04:57 iveqy joined #git
04:57 iveqy_ joined #git
04:58 freimatz joined #git
05:00 dermoth joined #git
05:01 _xor joined #git
05:04 SwiftMatt joined #git
05:06 ejnahc joined #git
05:08 SpeakerToMeat joined #git
05:12 svm_invictvs joined #git
05:14 justanotheruser joined #git
05:14 eizua joined #git
05:16 eizua joined #git
05:17 fuchstronaut joined #git
05:21 stuh84 joined #git
05:23 zeroed joined #git
05:24 Mowgli joined #git
05:24 mattcen joined #git
05:25 xameada joined #git
05:29 rominronin joined #git
05:32 LeyaL joined #git
05:34 fedepad joined #git
05:35 kyledr joined #git
05:36 sea-gull1 joined #git
05:36 overlord_tm joined #git
05:37 ewet joined #git
05:37 lordjancso joined #git
05:40 nidr0x joined #git
05:41 Kitty- joined #git
05:41 lindenle joined #git
05:50 vdv joined #git
05:52 Raging_Hog joined #git
05:54 Lipp joined #git
05:54 sea-gull_ joined #git
05:56 Anja joined #git
06:01 kyledr joined #git
06:01 mangolisk_ joined #git
06:03 cyan__ joined #git
06:04 LucaTM joined #git
06:04 Sleepy63 joined #git
06:06 lb1c joined #git
06:09 eizua joined #git
06:11 dreiss joined #git
06:11 eizua joined #git
06:14 timvisher joined #git
06:16 masuberu joined #git
06:16 freimatz joined #git
06:17 snthsnth joined #git
06:19 fuchstronaut joined #git
06:19 star_prone joined #git
06:20 thiago joined #git
06:21 ewet joined #git
06:21 pikajude joined #git
06:21 lmatteis joined #git
06:21 moop joined #git
06:21 iota_pi joined #git
06:22 happy-dude joined #git
06:22 JeroenT joined #git
06:22 duderonomy joined #git
06:23 dpower joined #git
06:23 Lor_ joined #git
06:24 mattcen joined #git
06:25 rominronin joined #git
06:25 Samual joined #git
06:26 yaewa joined #git
06:26 chele joined #git
06:27 NeverDie joined #git
06:27 xaa joined #git
06:28 roelmonnens joined #git
06:34 zacts joined #git
06:39 eduardas_m joined #git
06:40 qt-x joined #git
06:41 Alienpruts joined #git
06:41 maret joined #git
06:41 thierryp joined #git
06:42 rorro joined #git
06:42 Anja joined #git
06:42 thierryp joined #git
06:43 ThomasLocke_ joined #git
06:44 jnavila joined #git
06:45 elect joined #git
06:46 ahmed_elgabri joined #git
06:48 freimatz joined #git
06:48 Guest84613 joined #git
06:49 King_Hual joined #git
06:49 Guest84613 left #git
06:50 ogny joined #git
06:50 dpower joined #git
06:51 Lor_ joined #git
06:51 moop joined #git
06:51 happy-dude joined #git
06:51 iota_pi joined #git
06:52 lmatteis joined #git
06:53 submitnine joined #git
06:55 Endarked joined #git
06:58 rchavik joined #git
06:59 Anja joined #git
06:59 overlord_tm joined #git
07:00 blahdodo joined #git
07:02 Lipp joined #git
07:03 timvisher joined #git
07:03 MacWinne_ joined #git
07:03 dunaeth joined #git
07:04 cdown joined #git
07:05 thiago joined #git
07:06 Darcidride joined #git
07:10 eka joined #git
07:12 star_prone joined #git
07:12 TomyWork joined #git
07:12 Guest23964 joined #git
07:13 dc5ala joined #git
07:13 encod3 joined #git
07:16 Sceorem joined #git
07:18 apotry joined #git
07:19 laumann joined #git
07:19 fuchstronaut joined #git
07:21 watersoul joined #git
07:22 tzafrir joined #git
07:23 mattcen joined #git
07:25 fedepad joined #git
07:26 rchavik joined #git
07:27 encod3 joined #git
07:27 alcohol https://toggl.com/programming-princess
07:27 clmsy joined #git
07:28 Balliad joined #git
07:32 peaches joined #git
07:32 peaches left #git
07:33 matrixa1 joined #git
07:35 delboy1978uk joined #git
07:36 delboy1978uk fatal: Pathspec '/path/to/file' is in submodule 'path/to'
07:36 delboy1978uk how do i kill this submodule? i already deleted its .git folder
07:39 thiago git rm --cached path/to
07:39 thiago you need to remove the submodule's link from the parent repo first
07:39 delboy1978uk aahh! thanks thiago
07:40 ayjay_t joined #git
07:40 roelmonn_ joined #git
07:43 berndl joined #git
07:43 moei joined #git
07:45 delboy1978uk thiago: that diud the trick, thanks again for your help!
07:47 delboy1978uk left #git
07:52 Mobutils joined #git
07:54 marcogmonteiro joined #git
07:56 fatalhalt joined #git
07:58 compeman joined #git
08:00 barduck joined #git
08:01 acetakwas joined #git
08:03 jonky joined #git
08:03 neofelis joined #git
08:03 durham joined #git
08:04 jonky hi
08:06 jonky im having a strange problem. when pulling from upstream, git doesn't seem to be updating files
08:07 fuchstronaut joined #git
08:07 i7c Does it say anything?
08:07 elastix joined #git
08:08 jonky 'already up-to-date'
08:08 jknetl joined #git
08:09 MrWoohoo joined #git
08:09 jonky i've triple checked that im on the correct branch
08:09 berndl jonky: Is the upstream correct?
08:09 jonky berndl, do you mean the correct url?
08:09 i7c It seems the remote does not contain the commit you think it does.
08:10 berndl jonky: yes.
08:10 jonky berndl, yea, it's correct
08:10 i7c Or in other words, the remote has nothing new. It has all the commits you have locally.
08:10 Orion3k joined #git
08:10 jonky i7c, well, I can see there are changes that I dont have
08:10 jonky maybe i'll checkout a specific commit
08:11 i7c How do you see that?
08:11 jonky i7c, by viewing the repository
08:11 bongjovi joined #git
08:11 i7c Okay maybe you’re just pulling the wrong branch.
08:11 djb-irc joined #git
08:11 jonky i7c, already checked that too
08:11 i7c git pull <remote> <branch>
08:12 i7c jonky, you can also do git fetch <remote> and see all the remote/branch things in !git lol
08:12 berndl jonky: does it happen in a fresh clone of the repo?
08:12 i7c !lol
08:12 gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all
08:12 jonky berndl, no
08:12 ac3takwas joined #git
08:12 timvisher joined #git
08:12 Lipp joined #git
08:13 armyriad joined #git
08:13 karthikeayan joined #git
08:13 berndl jonky: that's good. Then use the fresh clone instead.
08:14 karthikeayan joined #git
08:15 tvw joined #git
08:15 rominronin joined #git
08:16 Snugglebash joined #git
08:16 sl33k joined #git
08:16 jonky berndl, any idea what might have happened?
08:17 karthik2 joined #git
08:17 cdown joined #git
08:18 ahmed_elgabri joined #git
08:20 h1d joined #git
08:20 EY___ joined #git
08:21 HTandE joined #git
08:22 HTandE left #git
08:22 Amid joined #git
08:24 Drzacek joined #git
08:25 ochorocho__ joined #git
08:25 Drzacek Hi. If I want the QtCreator to use c++14 standard, how do I tell him to do so? I don't see any place to put it in Run&Build or Compiler options (not even a field to place arguments)
08:29 rtm1046 joined #git
08:31 djb-irc joined #git
08:31 gingitsune joined #git
08:31 gingitsune How would on go about merging a bunch of repos of the essentially the same thing to one repo with branches as the old repos.
08:33 mischat joined #git
08:34 chll_ joined #git
08:36 Kitty- joined #git
08:36 osse gingitsune: Step 1: pick one repo as the "main" one. Step 2: fetch all other repos into this one. Step 3: ??? Step 4: profit
08:37 cbreak submodules!
08:37 osse Were they all independently inited?
08:37 cbreak well, I guess if they are the same thing, git merge will do.
08:38 qt-x joined #git
08:39 gingitsune osse: yes
08:39 osse gingitsune: And are they the same code base? Or different components of a larger whole?
08:41 gingitsune osse: mostly the former
08:41 gingitsune i have a bunch these repos i need to refactor
08:41 takwas joined #git
08:42 grawity joined #git
08:43 titankiller joined #git
08:43 lindenle joined #git
08:47 dodobrain joined #git
08:47 Snugglebash joined #git
08:48 star_prone joined #git
08:48 nikivi joined #git
08:49 neofelis hello! question: what do I do if I want my working directory to look like it looked five commits back, without losing any work?
08:49 tobiasvl neofelis: you have uncommited work?
08:49 tobiasvl stash it?
08:49 neofelis not currently
08:49 neofelis I have committed work
08:50 tobiasvl oh, well then you can just checkout the earlier commit
08:50 neofelis aha, that simple
08:50 canton7 beware that this will !detach your HEAD
08:50 * neofelis is new to git
08:50 gitinfo [!detached] A detached HEAD(aka "no branch") occurs when your HEAD does not point at a branch. New commits will NOT be added to any branch, and can easily be !lost. This can happen if you a) check out a tag, remote tracking branch, or SHA; or b) if you are in a submodule; or you are in the middle of a c) am or d) rebase that is stuck/conflicted. See !reattach
08:51 gingitsune Is it possible to realign the old master branch with a branch of a different name in a new origin
08:51 neofelis so just "git checkout HEAD^5" and then when I'm done looking at the files as they were "git checkout SHA_of_current_head"
08:51 gingitsune Not sure whats the best way to word that
08:51 snyp joined #git
08:51 tobiasvl neofelis: don't check out the SHA_of_current_head, check out the branch
08:52 neofelis right. thanks!
08:52 tobiasvl neofelis: checking out the SHA would still leave you detached
08:52 tobiasvl so if you're on master now, for example, "git checkout master^5" (or HEAD^5, same thing) and then afterwards "git checkout master"
08:53 snyp I use bitbucket. I pushed a commit. But I did a --amend on my local repository, and now the next git push doesn't work. This has happened to be before, and I was a newbie and some branching thing happened after it suggested that I actually pull from the repo and then commit the changes (or something). How do I do this cleanly, i.e the --amend reflected directly in bitbucket's repo too?
08:53 gingitsune As I asked before. I want to merge repos A, B and C to a repo ABC with respective branches A, B and C.
08:53 neofelis tobiasvl: yes, worked like a charm. thanks again
08:54 gingitsune I would like the wireup deployed clones of the repo to refer to the respective branches of the new master repo.
08:54 gingitsune Is it possible?
08:54 tobiasvl gingitsune: yes, it's possible. depends what you mean by "old master branch" though. but I assume you just want "git checkout master; git reset --hard remote/branch" – but beware that reset --hard will throw away what's currently on master, and any uncommited changes
08:55 tobiasvl oh, I wrote that while you wrote your last message. no idea what "wireup deployed clones" are but read !deploy
08:55 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.html
08:55 snyp Here's the error by the way - http://paste.debian.net/857758/
08:55 tobiasvl gingitsune: "realign"… maybe you want --set-upstream-to as well then
08:55 tobiasvl not sure
08:56 neofelis snyp: either git push -f, or git fetch & git rebase & git push
08:56 tobiasvl snyp: !force_push
08:56 gitinfo snyp: If you need to overwrite the history of a remote git repository (very bad idea, see !rewrite), you can do so with `git push --force-with-lease`. Note the remote server may reject this. See man git-config and search for receive.denyNonFastForwards. Best practice is for upstream servers to denyNonFastForwards.
08:56 tobiasvl snyp: !rewrite
08:56 gitinfo snyp: Rewriting public history is not recommended. Everyone who has pulled the old history will have to do work (and you'll have to tell them to), so it's infinitely better to just move on. If you must, you can use `git push --force-with-lease <remote> <branch>` to force (and the remote may reject that, anyway). See http://goo.gl/waqum
08:56 neofelis probably the latter.
08:57 tobiasvl probably – but if you amend a commit I assume you probably want that commit to replace the old one :) only snyp knows for sure
08:57 Drzacek left #git
08:57 tobiasvl snyp: if other people use the repo as well, force pushing is not recommended and rebase should be used. if you're the only person who uses it, it's probably fine. read above.
08:57 WayToDoor joined #git
08:58 snyp tobiasvl: yes, i just pushed, then realized a mistake. only i use this repo.
08:58 snyp tobiasvl: i see.
08:58 averell joined #git
08:58 tobiasvl OK, if only you use it, it's probably fine to do a push --force-with-lease, but read the above caveats.
08:58 snyp hm
09:01 Lipp joined #git
09:01 timvisher joined #git
09:02 satya4ever joined #git
09:05 EY__ joined #git
09:05 achadwick joined #git
09:08 roue joined #git
09:09 rexnsh joined #git
09:10 chid joined #git
09:12 thiago joined #git
09:12 re1 joined #git
09:13 DieguezZ joined #git
09:13 redeemed joined #git
09:15 Shentino joined #git
09:18 eka joined #git
09:19 a_thakur joined #git
09:20 Sadeukko joined #git
09:21 wrksx joined #git
09:21 wrksx I've been thinking some more about the simplistic deployment tool I want to setup using git
09:21 _ikke_ wrksx: git is not a deployment tool
09:22 ruskie joined #git
09:22 wrksx _ikke_, yeah been there yesterday mate
09:22 wrksx _ikke_, I'll wright a script to deploy, git happen to be the source container
09:22 wrksx wright/write
09:22 wrksx =) nice typo
09:22 _ikke_ Alright, sounds good
09:25 wrksx So the script is gonna take care of the source extract from git probably using archive, please advices if you have better ideas. Also, to be able to use git command on a git (bare) repo you need read permission on the repo, right?
09:26 wrksx (I meant better ideas than git-archive to extract sources)
09:28 thiago git archive is correct
09:31 karthik2 left #git
09:31 courrier joined #git
09:31 mischat joined #git
09:31 karthik2 joined #git
09:32 _ikke_ It's excelent even
09:35 osse gingitsune: adding B and C as remotes to A and then fetch from them achieves the initial goal
09:35 osse gingitsune: but it is a bit messy, and there is no silver bullet here
09:35 gingitsune osse: Thanks, I'm just gonna try and get my hands dirty
09:37 fuchstronaut joined #git
09:37 osse gingitsune: but if what you said is true then the branches will have no common commits and be entirely separate
09:37 osse ie. repo A will now have three roots (initial commits)
09:38 gingitsune osse: I assume thats not a hude drawback?
09:39 wrksx I've got trouble figuring out how to setup the whole thing actually. I'm using gitolite to host my repos, so all the git repos are owned by the gitolite user. I think I'd need to setup a specific user (deployer) to run the deploy script with specific permissions. I've always had trouble to wrap my head around user and permissions. For instance how to give my deployer user read access to the gitolite repos?
09:40 wrksx May not be the best place to talk about this I know
09:41 _ikke_ wrksx: The gitolite documentation is very good
09:41 _ikke_ wrksx: Just give it a specific ssh key, which only get read access
09:43 wrksx _ikke_, ah I think I got what you mean. Actually I was thinking about using git-archves on the gitolite repos directly, since they are on the same box.
09:43 gingitsune with git clone --mirror can i control wich branches get updated?
09:44 antlarr joined #git
09:44 wrksx _ikke_, But what I understand from what you said it might be a better idea to get another repo owned by my deployer user?
09:44 mattcen joined #git
09:44 wrksx but this won't use ssh because it's all local...
09:45 dan2k3k4 joined #git
09:48 wrksx Found a nice SO question about that matter, it says the easiest route would be to use ssh anyway, because gitolite's repos are not meant to be cloned locally
09:50 ferr joined #git
09:52 vdv joined #git
09:52 dwmw2_gone joined #git
09:53 thiago joined #git
09:54 dispix joined #git
09:56 guardianJ joined #git
09:58 testaaaaa joined #git
09:58 achadwick joined #git
09:58 xaa joined #git
10:01 a_thakur_ joined #git
10:03 dispix joined #git
10:04 m0viefreak joined #git
10:04 osse gingitsune: Its fine, git.git has seven
10:05 osse But they're merged into one, though
10:05 dispix Hey there !
10:05 gitinfo dispix: 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.
10:06 dispix Would anyone be available, just a couple of minutes, to help understand a weird rebase I did exploring git features ? :)
10:07 raven-au joined #git
10:08 tvw joined #git
10:08 roelmonnens joined #git
10:09 rorro joined #git
10:09 thiago ask your question, don't ask for a person
10:12 dispix Well i made a small drawing cause the visual might be easier to understand : https://goo.gl/hfReGy
10:12 dispix Basically, I don't understand the way my command (`git rebase HEAD~1^2`) behaved in this case.
10:13 dispix Why did it 'copied' commit F and C but not D and E for example ?
10:14 osse that looks strange to me. it should have also made D' and E'
10:15 star_prone joined #git
10:16 dispix Well, E is a merge commit between B and D so I guess there's no reason copying it ?
10:16 dispix I don't know about D though
10:16 thiago because E is a merge commit
10:16 osse yes that explains E
10:16 thiago Rule of thumb: don't rebase merges
10:17 osse D could also have been dropped if B introduce the same changes as D
10:17 testaaaaa joined #git
10:17 thiago If you think it's a good idea, you're probably wrong. If you think you need to do it, reconsider.
10:17 thiago it's possible, but as a general rule, avoid them
10:18 dispix It's actually a dummy project to see how git behave in edge cases
10:18 dispix So yeah, I may never have a use case for that type of rebase.
10:19 thiago I *think* rebase operates like this: if the other side of the merge that is being rebased was commits that weren't in the rebase itself, it re-does the merge
10:19 thiago otherwise, it linearises history
10:19 a_thakur joined #git
10:21 dispix I added the screenshots to the drawing, I think D may have been bypassed because it's similar to commit B
10:22 sostenuto joined #git
10:22 dispix But otherwise, good to know that you shouldn't rebase merge commits, that seems quite logical.
10:22 diogenese joined #git
10:25 greitas joined #git
10:26 ochorocho__ joined #git
10:26 fuchstronaut joined #git
10:27 star_prone joined #git
10:28 notebox joined #git
10:29 encod3 joined #git
10:30 star_prone joined #git
10:36 encod3 joined #git
10:36 cdown joined #git
10:36 dwmw2_gone joined #git
10:39 Snugglebash joined #git
10:40 dan2k3k4 joined #git
10:41 tvw joined #git
10:43 JSharpe joined #git
10:43 Mobutils joined #git
10:44 kurkale6ka joined #git
10:45 greendo joined #git
10:48 rorro joined #git
10:49 strk joined #git
10:49 strk how am I supposed to create new branches to a remote ? I'm getting: error: unable to push to unqualified destination: svn-2.3
10:49 strk from: git push gitlab 2.3:svn-2.3
10:51 canton7 strk, http://git.661346.n2.nabble.com/error-with-g​it-push-origin-HEAD-newbranch-td6333596.html
10:51 canton7 strk, git doesn't know whether you want a new branch or a new tag
10:51 canton7 strk, so you have to tell it with 'git push gitlab 2.3:refs/heads/svn-2.3' (or git push gitlab 2.3:refs/tags/svn-2.3')
10:53 strk thanks !
10:55 berndl joined #git
10:56 strk left #git
10:56 dwmw2_gone joined #git
10:58 tzafrir left #git
10:59 ToBeCloud joined #git
11:00 takwas joined #git
11:02 sonikspin joined #git
11:03 mischat joined #git
11:04 Lipp_ joined #git
11:04 Lipp joined #git
11:04 wololokatamari[m joined #git
11:05 afuentes joined #git
11:06 ocbtec joined #git
11:07 Snugglebash joined #git
11:08 raijin joined #git
11:08 kikijiki[m] joined #git
11:08 hallogallo[m] joined #git
11:08 Remramm joined #git
11:08 M-shine joined #git
11:08 Ineentho[m] joined #git
11:08 M-mistake joined #git
11:08 M-meznak joined #git
11:08 Lipp__ joined #git
11:08 aviraldg joined #git
11:10 Limes_ joined #git
11:12 rnsanchez joined #git
11:12 kjsage[m] joined #git
11:13 [cust] joined #git
11:13 dan2k3k4 joined #git
11:19 encod3 joined #git
11:20 Limes_ joined #git
11:22 rominronin joined #git
11:23 encod3_ joined #git
11:24 berndl joined #git
11:24 Murii joined #git
11:25 WayToDoor joined #git
11:29 ojdo joined #git
11:29 star_prone joined #git
11:29 dsdeiz joined #git
11:30 squirrel joined #git
11:31 Snugglebash joined #git
11:32 squirrel is this how i put a new branch from upstream into origin? fetch upstream; checkout new-branch; push origin new-branch; branch -u origin/new-branch
11:32 wrksx Am I correct to assume a bare repo cannot be on a specific branch, that a working tree feature right?
11:33 _ikke_ wrksx: It does have a HEAD (which decides what branch is 'current')
11:33 _ikke_ But that's more used when cloning to decide what branch gets checked out b y default
11:34 wrksx you mean checked out from the bare?
11:34 wrksx _ikke_, ^
11:34 nettoweb joined #git
11:34 wrksx cloning from the bare
11:34 i7c yup that he meant.
11:35 wrksx alright
11:35 wrksx am I also correct to assume one can git-archive from a bare?
11:35 Seveas correct
11:35 _ikke_ sure
11:35 nuck joined #git
11:36 jimi_ joined #git
11:36 wrksx I'm reading git-archive doc, and it looks like one can also git-acrchive from a remote directly
11:36 _ikke_ squirrel: simplest: git checkout -b new-branch; git push -u origin new-branch
11:36 _ikke_ yes
11:37 Seveas wrksx: only if the remote has enabled it (not the default)
11:37 mizu_no_oto joined #git
11:37 _ikke_ wrksx: that only applies to the git-daemon
11:37 mischat joined #git
11:37 Seveas _ikke_: pretty sure it works over smart http and ssh too
11:38 _ikke_ Seveas: I mean that it has to be enabled
11:38 _ikke_ on ssh it works by default, and I think on smart http(s) also
11:38 wrksx I'll give it a try then
11:38 Seveas _ikke_: oh, that's news to me
11:38 _ikke_ github does not offer it though (they provide their own tarballs)
11:39 encod3 joined #git
11:39 karthikeayan joined #git
11:39 rorro joined #git
11:39 Seveas fatal: Operation not supported by protocol.
11:40 Seveas (for https)
11:40 squirrel _ikke_: oh i see, thanks!
11:40 Seveas for ssh it does indeed work by default
11:40 _ikke_ ok
11:40 dwmw2_gone joined #git
11:40 Seveas TIL. Thanks _ikke_ :)
11:40 eizua joined #git
11:41 karthike1991 joined #git
11:43 _ikke_ Seveas: https://stackoverflow.com/a/11258694
11:43 _ikke_ seems to imply it should also work in smart-http
11:43 a_thakur joined #git
11:44 wrksx hum actually to have a remote I need a repo
11:45 sdothum joined #git
11:45 wrksx is this possible to git-arch from a distant repo without setting up a local repo?
11:45 Seveas well, that error came from a live smart http server. Though a somewhat special one
11:45 Seveas I do recall seeing comments in the source code about smart http not support ing git-archive though
11:46 Seveas yeah, http-backend.c has no support for upload-archive
11:47 wrksx Nm, it worked, just had a typo in the hostname :s
11:48 wrksx It feels so stupid to use ssh to reach a local repo
11:48 bremner stupider than http?
11:48 wrksx certainly not
11:48 Seveas why would you use ssh to reach a local repo?
11:49 wrksx Seveas, because the repo is hosted by gitolite
11:49 _ikke_ Seveas: That's a pitty
11:49 wrksx thus the filesystem permissions prevent me from accessing that
11:49 fuchstronaut joined #git
11:49 Seveas ah
11:49 bremner that's probably a good thing, if the repo is actually shared
11:51 _ikke_ yeah, and it lets you use gitolites acl
11:51 _ikke_ special deploy user that can only read for example
11:51 wrksx but forces me to use ssh to reach local repo
11:51 _ikke_ Should not be too big of a problem, right?
11:51 wrksx nope
11:51 wrksx 'perfectionism'
11:52 karthikeayan1991 joined #git
11:52 _ikke_ It's just a abstraction
11:53 wrksx hu what u mean
11:53 _ikke_ You get your git repositories from git@server:path, and don't care whether server is local or not
11:54 _ikke_ it simplifies authorization and authentication
11:54 wrksx right
11:54 cdown joined #git
11:55 wrksx movin on, if I use a tag as a target for my git-archive command, I do not need to specifiy any branch since a tag points to a commit anyway right?
11:55 _ikke_ right. git archive asks for a <treeish>, so anything that can be resolved to a tree
11:55 _ikke_ so both a branch and a tag are refs that point to a commit that points to a tree
11:55 fuchstronaut_ joined #git
11:56 _ikke_ Note that git archive behaves a bit different when you specify a tree or a commit
11:56 wrksx _ikke_, based on what you said pointing to a tag should trigger the tree like behavior, right?
11:57 wrksx pointing to/ specifying a tag as git arch targety
11:57 wrouesnel3 joined #git
11:57 _ikke_ "git archive behaves differently when given a tree ID versus when given a commit ID or tag ID. In the first case the current time is used as the modification time of each file in the archive. In the latter case the commit time as recorded in the referenced commit object is used instead."
11:58 Seveas _ikke_: just did some quick grepping: the remote-helper mechanism (and thus also remote-curl for http) does not support git-archive, so both client and server side are missing :)
11:58 timvisher joined #git
11:58 wrksx _that's a nice little feature
11:59 _ikke_ It's because trees don't have a date/time associated to it, so it cannot use it, but commits do
11:59 jas4711 joined #git
11:59 wrksx It's cool that it uses it
12:00 rpg joined #git
12:00 kyledr joined #git
12:00 fahadash joined #git
12:00 wrksx I just found out that it doesn't seem possible to use git-archive to output a directory as oposed to an archive. I guess I'll have to uncompress the archive after the export. Any suggestion welcome =)
12:01 Seveas wrksx: git archive | tar x :)
12:01 afuentes joined #git
12:01 canton7 wrksx, man git archive, see the first example
12:01 gitinfo wrksx: the git-archive manpage is available at http://jk.gs/git-archive.html
12:01 _ikke_ wrksx: git archive uses tar as a streaming format
12:01 Seveas git-archive's --prefix parameter may be of use to you as well
12:01 _ikke_ yeah, if you want the files to live in a directory
12:02 _ikke_ note that you want to add the / suffix
12:02 nettoweb joined #git
12:02 _ikke_ otherwise it would just prepend the prefix to the filenames (happened to me once :P)
12:02 _ikke_ --prefix=dir/
12:02 wrksx right got it thx for the tip
12:02 Seveas happened to me more than once :)
12:02 _ikke_ haha
12:05 wrksx guys that works so fuckin right
12:06 _ikke_ Why do you think we are so in favor if git-archive?
12:06 jaguarmagenta joined #git
12:06 _ikke_ sorry for the rhetorical questions :-)
12:06 wrksx now I know
12:07 mischat joined #git
12:07 wrksx specially the --remote=git@server:repo <tag> which prevents me from duplicating every repo
12:07 wrksx I'm fuckin pleased
12:08 synthroid joined #git
12:08 dispix joined #git
12:09 johnmilton joined #git
12:09 no_gravity joined #git
12:09 no_gravity I'm having a strange issue with git. When I push it to github it says "Everything up-to-date". But it's not.
12:09 no_gravity Shouldnt this update it all? git add -A; git commit -m "bla bla";git push ...
12:10 _ikke_ no_gravity: depends on what you specify with git push and your push.default config setting
12:10 SwiftMatt joined #git
12:11 _ikke_ no_gravity: when git says Everything up-to-date, it means that the remote has everything it's trying to push
12:11 _ikke_ So if that doesn't match you expectation, it probably means it isn't pushing what you think
12:13 no_gravity _ikke_: Well, I want to push everything :)
12:13 _ikke_ no_gravity: Do you want to push *all* branches?
12:13 _ikke_ Or just the current branch you are working on?
12:14 ayjay_t joined #git
12:14 no_gravity _ikke_: There is only one branch I think.
12:14 _ikke_ no_gravity: what does git branch -vv return?
12:14 flaviodesousa joined #git
12:15 no_gravity * master ... [origin/master: ahead 37]
12:16 _ikke_ Ok, next step: git config push.default
12:16 no_gravity ok
12:16 barajasfab joined #git
12:16 no_gravity and then?
12:16 _ikke_ what does that return?
12:16 no_gravity Nothing
12:17 LBV_User joined #git
12:17 _ikke_ ok, and git version?
12:17 no_gravity git version 2.7.4
12:18 bruce_lee joined #git
12:18 bruce_lee joined #git
12:18 _ikke_ ok. `git push`
12:18 _ikke_ what does that return?
12:20 sl33k_ joined #git
12:20 nowox joined #git
12:21 buffal0 joined #git
12:22 dwmw2_gone joined #git
12:24 mischat joined #git
12:24 no_gravity asks for username
12:24 wrksx Would there be a way to list all tags contained on a specific branch?
12:25 wrksx on a remote repo using git@host:repo
12:26 dwmw2_gone joined #git
12:27 ochorocho__ joined #git
12:27 dave0x6d joined #git
12:29 rtm1046 joined #git
12:32 jeffreylevesque joined #git
12:33 no_gravity Shouldn't git tell me when I push that it holds stuff back? Instead of just saying "Everything up-to-date", which is not true?
12:33 no_gravity There are files I changed minutes ago. They *must* be newer here then on GitHub.
12:33 UncleCJ joined #git
12:33 _ikke_ no_gravity: git never holds stuff back
12:34 no_gravity _ikke_: So why does it tell me stuff is up-to-date when I just changed stuff?
12:34 _ikke_ wrksx: You can get a list of all tags, but not any reachability info remotely
12:35 _ikke_ no_gravity: try git push origin master
12:35 no_gravity _ikke_: "Everything up-to-date"
12:35 gingitsune Can I push from one bare repo straight into another?
12:35 _ikke_ gingitsune: sure
12:36 _ikke_ no_gravity: git fetch, and then git branch -vv again
12:36 _ikke_ git fetch origin
12:36 no_gravity _ikke_: What does that do?
12:37 DeltaHeavy joined #git
12:37 _ikke_ get everything from the remote repository, and update your local information about it
12:37 gingitsune _ikke_: it might be worth noting that they are different repos
12:37 gingitsune as in they have been inited seporately
12:37 gingitsune the code is related but there is no common history
12:38 _ikke_ gingitsune: git doesn't care, as long as you don't push to an existing branch
12:38 gingitsune the branch doesn't exist i got "git push remote master:new_branch"
12:39 dwmw2_gone joined #git
12:39 _ikke_ git push remote master:refs/heads/new_branch
12:40 gingitsune _ikke_: whats the refs/heads/ bussiness about?
12:40 _ikke_ gingitsune: fully qualified ref name
12:41 gingitsune I feel the need to read a git book
12:41 _ikke_ branches in git always start with refs/heads, but that is usually hidden
12:41 _ikke_ but if you push using a refspec to a new branch, git doesn't know what you want, so you have to be explicit
12:42 UncleCJ Does anyone else have the impression that git knowledge gets buried in syntax, with little to no references to sensible articles like this? https://sethrobertson.github.io/GitBestPractices/
12:43 eroussel joined #git
12:43 iota_pi joined #git
12:43 _ikke_ UncleCJ: too bad SethRobertson is not 'active' anymore to update references
12:43 _ikke_ I haven't seen him since a few years anyway
12:44 CEnnis91 joined #git
12:44 _ikke_ UncleCJ: We do have a trigger for that one: !best_practices
12:44 gitinfo UncleCJ: There is no one right answer for git best practices, but a consensus from #git is available at http://sethrobertson.github.com/GitBestPractices/
12:44 lmatteis joined #git
12:45 UncleCJ _ikke_: Yeah... wasn't aware of that. Anyway it's a matter of approach, so many poor git articles...
12:45 barteks2x joined #git
12:45 _ikke_ yes, they are plentyful
12:45 UncleCJ _ikke_: Nice trigger! :-P
12:46 _ikke_ Hard to get rid of poor git articles
12:46 gingitsune _ikke_: I'm somewhat stumped. It gives me everything up to date. But if i do git branch -v the branch is not there
12:46 lindenle joined #git
12:46 _ikke_ gingitsune: git branch -r
12:47 no_gravity _ikke_: I figured it out! On this machine, git seems to not commit. It just blurbs out "Please tell me who you are"
12:47 no_gravity _ikke_: After I handed over an "email" and a "name" git pushes now :)
12:47 _ikke_ no_gravity: Good, it's always good to read the messages git gives you
12:48 greitas left #git
12:48 gingitsune _ikke_: git branch -r output remote/branch_name in the bare repo i pushd from. But i see nothing in the repo it supposedly pushed to
12:48 dwmw2_gone joined #git
12:48 dendazen joined #git
12:49 brokensyntax joined #git
12:49 _ikke_ gingitsune: and you verified the remote url?
12:49 _ikke_ git remote -v
12:49 gingitsune _ikke_: ye they are both on a machine i control
12:50 gingitsune _ikke_: owh boy
12:50 gingitsune jeez. not enough '../'
12:50 gingitsune Thanks :)
12:50 _ikke_ :-)
12:51 _ikke_ One thing you learn here is never to trust any assumptions :P
12:52 wrksx git ls-remote returns refs with some stuff I'd like to get rid of. Actually I'd like to only grab the tag name itself
12:52 anonymuse joined #git
12:52 wrksx anyway to do that?
12:52 gingitsune _ikke_: out of curiosity i tried pushing without ref/heads and it worked just fine
12:52 _ikke_ gingitsune: Note sure under what condition git requires you to be explicit
12:53 _ikke_ wrksx: git rev-parse and git name-rev can be used to transform ref names
12:53 fuchstronaut joined #git
12:54 wrksx _ikke_, okay thank you will have a look
12:54 dendazen joined #git
12:54 durham joined #git
12:54 _ikke_ wrksx: ah, that does not work with remote refs you don't have
12:55 wrksx _ikke_, exactly what I was investigating
12:56 mischat_ joined #git
12:56 _ikke_ wrksx: git ls-remote origin | cut -f2 | grep tags | grep -v "\^" | sed 's|refs/tags/||'
12:57 harish joined #git
12:57 ash_workz joined #git
12:58 wrksx git ls-remote --tags <remote> | awk '{sub(/refs\/tags\//, "", $2); print $2}'
12:58 raven-au joined #git
12:58 _ikke_ hehe
12:58 wrksx _ikke_, with some help from #awk =)
12:58 _ikke_ I recently learned a bit more about awk, but it's not in my active toolset yet :P
12:59 wrksx I found it really really nice
12:59 _ikke_ me too
12:59 _ikke_ It can even be used as a calculator :P
12:59 wrksx Yeah so many stuff that it can do
12:59 wrksx Amazing
13:00 timvisher joined #git
13:02 red82 joined #git
13:03 jwest joined #git
13:04 FirewalkR joined #git
13:05 MillerBoss joined #git
13:06 faux___ joined #git
13:06 aswen joined #git
13:06 harish joined #git
13:06 GodGinrai joined #git
13:08 no_gravity left #git
13:09 mtutty joined #git
13:12 takwas joined #git
13:12 wrksx How could I make sure a given tag is contained on a given branch?
13:13 wrksx but on a remote repo
13:13 t2mkn joined #git
13:13 mtutty Can anybody explain why I can't do git worktree add /some/path master on a bare repo?
13:16 blackwind_123 joined #git
13:16 _ikke_ wrksx: you can't, you don't have enough information
13:16 _ikke_ the git protocol does not allow you to ask such questions
13:17 wrksx _ikke_, oh I'm screwed
13:17 rorro joined #git
13:17 _ikke_ What do you want to achieve?
13:17 wrksx cause I want to use tags to mark deployable points
13:17 wrksx so I want a script that lists all deployable points
13:17 MrRoot joined #git
13:18 dwmw2_gone joined #git
13:18 Ediz joined #git
13:18 wrksx but I'd also like not all tags to show up, only those of the deploy branch
13:18 ShekharReddy joined #git
13:18 wrksx does that make sense
13:19 _ikke_ Can't you namespace 'deploy' tags?
13:19 _ikke_ refs/tags/deploy/...
13:19 wrksx Probably could.
13:19 wrksx What would be nice also is that not anyone could tag something ready for deploymnt
13:20 wrouesnel1 joined #git
13:20 _ikke_ not anyone or not everyone?
13:20 wrksx everyone sorry
13:20 _ikke_ You can use gitolite that prevent people from pushing tags that match a certain pattern
13:20 wrksx _ikke_, hu! I assumed that gitolite couldn't be so fine grained. Looks like I was totally wrong
13:21 wrksx _ikke_, thanks for the help mate
13:21 wrksx that sounds like a great solution
13:21 _ikke_ gitolite is really sophisticated
13:23 Kitty- joined #git
13:23 JamesBaxter joined #git
13:25 converge joined #git
13:25 jaafar joined #git
13:26 shmup joined #git
13:27 jimi_ joined #git
13:27 shmup curious about gitconfig email. at work it's configured for work related pushes. i do a few side project stuff here, and push out to my github repositories. i was kind of hoping there'd be a way to configure my email under the [Github] block for Github remote related pushes. is this a thing?
13:28 shmup or is there some popular technique for handling multiple git config email addresses? maybe local git config/project based settings? anything "better"?
13:28 gingitsune Is it possible to use 'git push origin' ommitting the branch?
13:28 rgrinberg joined #git
13:28 tobiasvl shmup: yeah, I just don
13:28 gingitsune Would git know which remote branch is meant?
13:28 tobiasvl don't have a global email set up
13:28 selckin gingitsune: tracking branch
13:28 tobiasvl so I need to fix it locally in every repo. that way I know it's always correct
13:29 shmup tobiasvl: oh, cool. so by removing the global email it won't even allow a repo to push? that's nice
13:29 shmup if so. meaning it will force local repo configuration
13:29 tobiasvl shmup: well, it will allow you to commit, but it will warn you when you do
13:29 selckin email is stored on commit not push
13:29 shmup oh. all right
13:29 tobiasvl shmup: it will give the e-mail as username@hostname or something, and warn
13:30 ashnur i can not use git properly from windows because despite all attempts, it doesn't uses my ssh keys :(
13:30 alex1a joined #git
13:30 selckin ashnur: what have you tried, and which git
13:30 tobiasvl ashnur: do the ssh keys work for other stuff?
13:30 shmup tobiasvl: ah, noted. thanks
13:30 ivan_on_trac joined #git
13:30 ashnur git version 2.8.3.windows.1
13:31 selckin set GIT_SSH to plink and add keys to pagent (all putty)
13:31 ivan_on_trac hi
13:31 ivan_on_trac How do I calculate the number of lines changed between branches?
13:31 ashnur I can use the ssh keys with the ssh command and they work
13:32 mlq joined #git
13:32 trondth what is the best way to create a pull-request on a github repo, where you do not have the right to push?
13:33 selckin fork it
13:33 trondth thankyou.
13:33 GenZai joined #git
13:33 mlq Hi! I have a question: Is it possible to checkout a whole directory of a different commit into a new directory? With git show, I could do that for a single file and then just pipe the output to a new file. What I want to do is to apply latexdiff between a previous commit and the latest one automatically
13:33 dviola joined #git
13:36 selckin git checkout sha -- path
13:36 t2mkn left #git
13:37 shmup left #git
13:37 anonymuse joined #git
13:39 mtutty mlq: You can use git diff for that.  You can specify from and to commit hashes, use the "
13:39 mtutty ".." or "..." notation, and then use -- <path> for the specific file you're looking for.
13:40 fuchstronaut joined #git
13:42 Emperor_Earth joined #git
13:42 ExoUNX joined #git
13:42 mlq mtutty: The problem is that I do not want to checkout a single file, since the document consists out of many different files and thus, I need to flatten the tex file in the end. Therefor I would like to checkout the whole directory ./doc of commit 12345 into ./doc-old for instance
13:44 timvisher joined #git
13:45 sangy joined #git
13:47 mlq Ah, I can use git archive and then --strip-components when unpacking :)
13:47 harish joined #git
13:48 |ifei5g00d joined #git
13:50 wgrant joined #git
13:51 elect_ joined #git
13:51 dwmw2_gone joined #git
13:51 gingitsune I have an idea that i want to have a master branch and then a prod branch for each client we have. When publishing a new feature we would merge from the master to the *client*-prod.
13:51 psprint joined #git
13:52 gingitsune Would there be any pitfalls to this sort of approach?
13:52 selckin !deploy
13:52 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.html
13:53 jantje joined #git
13:53 mtutty mlq: You can also diff the entire directory if you want, just omit the "-- <path>" element.
13:54 psprint Hello, 'git show SHA' shows commit's diff and standard git log header. I wonder: Can I make the header contain no date, the commit initial description contain list of touched files, the diff be more compact without dif --git a/ b/, index sha..sha, --- a/... +++b/, also no @@ -128,6 +128,16 @@ etc. ?
13:54 mtutty mlq: It's possible to use git worktree (git 2.5+) to materialize multiple branches into separate directories
13:55 jaequery joined #git
13:55 osse psprint: yes. use --pretty/--format. documented in git log
13:55 tristanp joined #git
13:56 mlq mtutty: I think worktrees could work, a simple diff won't because I need the latexdiff tool to create the actual diff. However, I've managed to build a solution with git archive :) Thanks for your help!
13:58 ivan_on_trac How to I count the lines changed between branches?
13:59 _ikke_ ivan_on_trac: git diff --state
13:59 mtutty psprint: You can use the --format option with a huge variety of special formatting strings: https://git-scm.com/docs/pretty-formats
13:59 _ikke_ --stat
13:59 _ikke_ ivan_on_trac: man git diff has more --*stat* options
13:59 gitinfo ivan_on_trac: the git-diff manpage is available at http://jk.gs/git-diff.html
14:00 kyledr joined #git
14:01 psprint mtutty: but it looks like removing things from diff's header isn't there
14:01 fmcgeough joined #git
14:01 thiago joined #git
14:02 thiago_ joined #git
14:04 fmcgeough joined #git
14:06 guampa joined #git
14:06 jaguarmagenta joined #git
14:07 Guest91461 joined #git
14:08 glen left #git
14:08 glen joined #git
14:08 glen hey. i'm using some python tool (agito) to svn2git convert. and i noticed it creates "tree" objects which are really empty (no files in). what to think of that, and is there (easy way) to cleanup that?
14:09 srg joined #git
14:10 srg Is it possible to have a git repo at /path/ (with files, directories, etc) and then another separate git repo at /path/site2/   (also with files, directories, etc)....without it being in the first repo? Should I add "site2/" to my parent repo's gitignore?
14:10 perlpilot srg: yes, it's possible.  If you really want to do that, an ignore couldn't hurt
14:11 nikivi joined #git
14:11 Snugglebash joined #git
14:12 treehug88 joined #git
14:13 hurikhan|Work joined #git
14:14 glen i created bugreport to agito which explains the situation: https://github.com/fragglet/agito/issues/14
14:15 matsaman joined #git
14:15 yqt joined #git
14:16 glen i'm not sure what to think of that. is such tree okay? or rather not? which part should be fixed?
14:16 eivarv joined #git
14:16 ivan_on_trac joined #git
14:17 AaronMT joined #git
14:17 anonymuse joined #git
14:18 axelitus joined #git
14:20 crose joined #git
14:20 govg joined #git
14:20 chrisgeorge_ joined #git
14:20 cagmz joined #git
14:22 al-damiri joined #git
14:23 jantje joined #git
14:25 mischat joined #git
14:25 _milky_ joined #git
14:25 jaguarmagenta joined #git
14:26 phaleth joined #git
14:27 kulelu88 joined #git
14:27 e14 joined #git
14:27 patarr joined #git
14:28 titankiller joined #git
14:29 Vampire0 glen, yes, empty commits are fine for git. they are maybe not nice, but technically ok. And yes, you can probably remove them. But maybe you should think about !svn2git
14:29 gitinfo glen: For a one-time conversion of SVN to git, svn2git should be used instead of git-svn. There are many tools with this name. Probably the best is the KDE one at https://github.com/svn-all-fast-export/svn2git. To analyze the SVN repo history upfront for building up proper rules for svn2git, you might use svneverever from here: http://blog.hartwork.org/?p=763
14:31 mtutty Can anybody explain why I can't do git worktree add /some/path master on a bare repo?
14:32 shinnya joined #git
14:33 DJ_Spies joined #git
14:36 raspado joined #git
14:36 synthroid joined #git
14:38 leeN joined #git
14:38 psprint looked at docs and it seems that it is impossible to show diff without header?
14:41 glen Vampire0: it's not empty commit i'm talking about, but tree objects that have nothing underneath them
14:41 shgysk8zer0 joined #git
14:43 glen Vampire0: not that matters, but i use agito because svn2git has it's own issues that agito solves. so please do not promote other tool. in fact i did not even ask that. please read again actual question
14:43 canuk joined #git
14:43 glen Vampire0: and "probably remove" i assued, that's why i asked "how"
14:43 osse psprint: using GIT_EXTERNAL_DIFF you can do whatever you want. eg. use regular diff or something else
14:45 DevAntoine joined #git
14:45 DevAntoine hi
14:45 grift joined #git
14:46 DevAntoine I've merge a branch FOO into develop. A given file add lot of change on FOO. After the merge all the added function on this file on FOO or not in this same file on develop. I had no conflict. WTF?!
14:46 tyreld_ joined #git
14:47 DevAntoine ok, git fetch doesn't pull the branch, I had to 1) git fetch 2) git checkout FOO 3) git pull 4) git checkout develop 5) git merge FOO
14:47 DevAntoine tedious
14:47 rpg joined #git
14:48 Shock_ joined #git
14:49 robotroll joined #git
14:50 Vampire0 glen, ah, if you didn't mean empty commit, I don't know. I know that you didn't ask for an alternate tool, I'm not legasthenic. Yet why should that prevent me from suggesting you to use it? o_O   I'm interested in what svn2git issues you see that agito solves. Of course any piece of software has its pros and cons. I'm really interested in what issues you had that agito solves.
14:52 re1 joined #git
14:53 nettoweb joined #git
14:56 xameada joined #git
14:57 rpg left #git
14:57 psprint osse: I would rather go for following zsh code, as GIT_DIFF_OPTS also fail (man diff doesn't reveal option to suppress header...): git show --word-diff=color  dd51d8a | (){ while read line; do [[ "$line" != (#s)($'\033'?##)(diff|index|---|+++)* ]] && print -r -- "$line"; done }
14:57 guampa joined #git
14:58 psprint but the escape codes should be catched better, and also hunk @@ lines
14:58 psprint hm and (){ isn't needed actually
14:59 srg left #git
15:00 grift joined #git
15:00 joki joined #git
15:02 einnjo joined #git
15:03 gemie joined #git
15:03 ghostlight joined #git
15:04 psprint (#s)($'\033'[^m]#m)#(diff|index|---|+++)* ]] quite good escape matching, ESC and then not-m, then m
15:04 gemie Hi, I have a git rebase interactive question. Let's say there is a branch called Foo. I have no commits on that branch. I then do a new branch called Bar from Foo. When doing git rebase -i <BASE>, should base be the last commit of Foo or my first commit on Bar?
15:04 gemie I want to squish all my commits on Bar to 1 commit.
15:05 dreiss joined #git
15:06 grift joined #git
15:06 astrofog joined #git
15:06 tobie joined #git
15:06 fuchstronaut How do I reset branch 'feature' back to 'origin/feature'. I accidentally rebased my local branch onto master before pulling changes from 'origin/master' to 'master'
15:07 harish_ joined #git
15:08 nettoweb joined #git
15:08 evanc joined #git
15:09 star_prone joined #git
15:09 chrisgeorge joined #git
15:09 zero7 joined #git
15:10 trig-ger joined #git
15:11 yena joined #git
15:12 overlord_tm joined #git
15:14 CheckDavid joined #git
15:14 lxnch joined #git
15:15 elijah joined #git
15:15 noahsussman joined #git
15:15 moritz fuchstronaut: when you're on branch feature, git reset origin/feature
15:16 moritz possibly with --hard
15:17 dwigton_ left #git
15:17 fuchstronaut thanks
15:18 fuchstronaut should I force-push my branch back to origin after rebase?
15:18 mischat joined #git
15:18 moritz if nobody has pulled that branch, that's a viable option
15:18 mischat joined #git
15:19 prosody joined #git
15:20 theresajayne joined #git
15:20 imanc joined #git
15:20 WillMoogle joined #git
15:21 MLM joined #git
15:23 shennyg joined #git
15:23 da5id joined #git
15:25 amaged__ joined #git
15:25 jumpman joined #git
15:29 Bengi_ joined #git
15:29 _Vi joined #git
15:31 iamdevnul joined #git
15:31 _milky_ left #git
15:31 _milky_ joined #git
15:32 JackMc joined #git
15:32 worstadmin joined #git
15:33 kutenai joined #git
15:33 salamanderrake joined #git
15:33 InfoTest joined #git
15:35 flupke joined #git
15:35 Darren_ joined #git
15:35 Tee_Pee joined #git
15:36 davisonio joined #git
15:36 beckyconning_ joined #git
15:36 kireevco_ joined #git
15:36 GameGamer43 joined #git
15:36 epic joined #git
15:36 boxrick1 joined #git
15:36 daxroc joined #git
15:36 DuncanT joined #git
15:36 discopatrick joined #git
15:36 spr0tsuki joined #git
15:36 ionelmc joined #git
15:36 addshore joined #git
15:38 mustmodify joined #git
15:39 glen Vampire0: i just realized whatevr the fix is to remove empty tree objects would involve rewriting history. so it's not good option as i have commit messages referring to other commits via sha1 hashes and git notes involved as well
15:40 parliament joined #git
15:40 mustmodify ok! I'm doing a live-coding presentation. But because I'm a loser, I worry I'm going to be like, "Oh there's a bug! I can't find it! There went 20 minutes!" So my big plan is to have tags at each step so I can just say "Well, whatever, (A) show me what's different and (b) no matter how much I've Fd up, just get rid of extra files, and no matter what, the files should look like tag step_5."
15:40 glen Vampire0: the basic principle it does migrate svn history better that svn2git and git-svn fail https://github.com/fraggle​t/agito/blob/master/README
15:40 ggherdov joined #git
15:40 ochorocho__ joined #git
15:40 mustmodify So what's the command for that? reset --hard?
15:40 mustmodify no...
15:40 mustmodify hm...
15:40 mustmodify git checkout tag
15:41 Mixologic_ joined #git
15:41 _ikke_ git checkout -f tag && git clean -f
15:41 mustmodify and then presumably some kind of flag for "get rid of files, ignore conflicts, etc."
15:41 synthroid joined #git
15:41 glen Vampire0: also i can customize in python code different parent commit to handle svn weird moves as git merge commits
15:41 mustmodify is there anything that can stop `git checkout -f tag` from working?
15:41 mustmodify conflicts, failed merges, ...
15:42 shgysk8zer0 joined #git
15:42 _ikke_ Don't think so, but doing a git reset should get rid of all the state files
15:42 jleclanche joined #git
15:42 glen mustmodify: yes, object that used to be file is locally as directory or vice versa if tree is not clean
15:42 _ikke_ so git reset && git checkout -f tag && git clean -f
15:43 glen mustmodify: i would add stash command to the pipeline: git stash && git checkout -f tag
15:43 mustmodify glen: so start with git clean -f?
15:43 _ikke_ clean -f will only apply to untracked files
15:43 glen mustmodify: and add --include-untracked to stash
15:43 mustmodify glen: Ah, that's what I was worrying about.
15:43 mustmodify ok.
15:43 glen mustmodify: so, this should work: git stash --include-untracked && git checkout -f tag
15:43 glen mustmodify: looks like you're trying to implent git-fire :D
15:43 mustmodify git stash --include-untracked && git checkout -f tag && git clean -f
15:43 glen https://github.com/qw3rtman/git-fire
15:44 glen mustmodify: you do not need git clean afaik because stash --include-untracked would clean them as well
15:44 glen but if you must, git clean -fdx perhaps even
15:44 mustmodify makes sense that git stash would do that.
15:45 mustmodify git stash --include-untracked && git checkout -f tag
15:45 littleforest joined #git
15:45 glen but give it a test! try to create scenario as well where some pathname becames from file to dir (and vice versa)
15:45 mustmodify I think my plan is to use sqlite3 because then my database will transition along with my examples.
15:45 mustmodify There will definitely be some testing. :P
15:46 mustmodify Thanks, all!
15:46 ss22ever22 joined #git
15:46 Mindfart joined #git
15:46 mustmodify "Disclaimer: Code can be re-written, but humans cannot."
15:47 ss22ever22 Hello can anybody help me in git cloning a website that has been blocked by a proxy server
15:47 mustmodify lol
15:47 Shentino joined #git
15:47 mustmodify you want to clone a website?
15:47 mustmodify hunh...
15:47 mustmodify do you have access to the server?
15:47 ss22ever22 No no from the website where I want to clone has been blocked
15:48 mustmodify Is this your website or someone else's?
15:48 ss22ever22 The college campus server has blocked it
15:48 ss22ever22 Someone else's
15:48 mustmodify I see.
15:49 dave0x6d joined #git
15:49 Lipp joined #git
15:49 mustmodify This 'website' ... do you mean you want to scrape a website and add it to git? Or that someone has a remote repo that you want to clone?
15:49 ss22ever22 Yes what can be the solution to clone it
15:50 canton7 https://www.gnu.org/software/wget/m​anual/wget.html#Recursive-Download ?
15:50 canton7 although of course you're not cloning any of the server-side stuff
15:51 govg joined #git
15:51 canton7 ... not a git question, though
15:51 mustmodify looks like ss22... is gone
15:51 bsanford joined #git
15:52 matsaman joined #git
15:52 Electrometro joined #git
15:54 le_melomane joined #git
15:58 EvilPenguin joined #git
15:59 bremner if I don't have a reflog, whats the most convenient way to approximate "git checkout master@{2016-05-01} ?
16:00 kyledr joined #git
16:01 jaequery joined #git
16:01 bremner some futzing with git log --until works, but I wondered if there was another option
16:01 pragmatism left #git
16:05 lindenle joined #git
16:06 laumann joined #git
16:07 notebox joined #git
16:08 e14 joined #git
16:09 overlord_tm joined #git
16:09 Oebele joined #git
16:10 notebox joined #git
16:11 jozwior joined #git
16:12 squirrel left #git
16:12 thiago joined #git
16:13 Seveas not that I'm aware of
16:16 red82 joined #git
16:18 Lipp joined #git
16:19 svm_invictvs joined #git
16:20 swalladge joined #git
16:20 matsaman .gitignore doesn't hide modified things from 'git status'?
16:20 rpg joined #git
16:20 drizbit joined #git
16:21 ilmari gitignore only affects what untracked files git cares about. once it's tracked it won't ignore it
16:21 skalpin not if they were previously tracked
16:22 pragmatism joined #git
16:22 fees joined #git
16:22 pragmatism left #git
16:22 thiago joined #git
16:23 oskarkv joined #git
16:23 ansimation joined #git
16:23 matsaman okay, so how can I fix it?
16:23 matsaman if they were previously tracked
16:24 skalpin if you want git to assume the file is unchanged you can 'git update-index --assume-unchanged <filepath>'
16:24 peaches joined #git
16:24 matsaman I want git status to never talk about it, ever, as if it had always been in .gitignore
16:24 peaches joined #git
16:25 skalpin http://stackoverflow.com/questions/936249/stop​-tracking-and-ignore-changes-to-a-file-in-git
16:26 govg joined #git
16:27 jaguarmagenta joined #git
16:27 matsaman thanks
16:29 harish joined #git
16:29 salamanderrake joined #git
16:29 axisys joined #git
16:30 Balliad joined #git
16:32 thiago joined #git
16:38 satya4ever joined #git
16:39 Guest23964 joined #git
16:40 spudowiar joined #git
16:40 nitric joined #git
16:43 Ldip joined #git
16:43 titankiller joined #git
16:44 Guest23964 left #git
16:45 romerocesar joined #git
16:46 masuberu joined #git
16:47 thierryp joined #git
16:49 ss22ever22 joined #git
16:49 dwmw2_gone joined #git
16:49 nettoweb joined #git
16:49 converge joined #git
16:49 converge joined #git
16:53 eroussel joined #git
16:53 Dazzyp joined #git
16:53 jnavila joined #git
16:57 madewokherd joined #git
16:58 User458764 joined #git
16:59 Doginal joined #git
16:59 aarobc joined #git
17:00 diogenese joined #git
17:01 Vampire0 glen, yes, git-svn produces sub-optimal history, as does the nirvdrum svn2git which is just a wrapper around git-svn and does some afterwards polishing. But that just shows that you didn't read my message at all. The !svn2git the snippet recommends is not the nirvdrum one, but the KDE one and you will hardly find any deficiencies with it. At least I didn't find some yet I was not able to easily overcome an
17:01 gitinfo glen: For a one-time conversion of SVN to git, svn2git should be used instead of git-svn. There are many tools with this name. Probably the best is the KDE one at https://github.com/svn-all-fast-export/svn2git. To analyze the SVN repo history upfront for building up proper rules for svn2git, you might use svneverever from here: http://blog.hartwork.org/?p=763
17:01 Vampire0 d I migrated many SVN repos to Git properly with it already, amongst them projects with really f***ed up history in SVN.
17:01 chrisgeorge_ joined #git
17:01 dreiss joined #git
17:03 SteffanW joined #git
17:04 EvilPenguin joined #git
17:07 cdesai joined #git
17:07 shgysk8zer0 joined #git
17:07 abh joined #git
17:07 ochorocho__ joined #git
17:07 jokester joined #git
17:08 red82 joined #git
17:08 takwas joined #git
17:08 asdofindia joined #git
17:08 e14 joined #git
17:08 sns joined #git
17:10 le_melomane joined #git
17:10 brokensyntax joined #git
17:10 mutex joined #git
17:11 saml I fast forwarded a merge. i want to revert and re-merge with --no-ff.  things are already pushed out.  what's a sane way to do this?
17:12 saml https://www.kernel.org/pub/software/scm/​git/docs/howto/revert-a-faulty-merge.txt
17:13 LeBlaaanc joined #git
17:13 dan2k3k4 joined #git
17:14 mustmodify left #git
17:16 freimatz joined #git
17:18 DevAntoine joined #git
17:18 brokensyntax joined #git
17:23 brokensyntax joined #git
17:23 fuchstronaut joined #git
17:23 gingitsune Would it be possible make a new branch from a bare repo?
17:24 danslo_ joined #git
17:24 cdg joined #git
17:25 nitric joined #git
17:25 synthroid joined #git
17:26 kerrick joined #git
17:26 Limes_ joined #git
17:27 saml git revert foo;  git checkout feature;  git rebase master;    git checkout master; git merge --no-ff feature;
17:28 Juzzika joined #git
17:28 fatalhalt joined #git
17:28 saml ah i had to revert the revert
17:30 swalladge joined #git
17:31 Remramm joined #git
17:31 Remramm joined #git
17:32 jost joined #git
17:33 blackwind_123 joined #git
17:34 dodobrain joined #git
17:36 govg joined #git
17:37 cagedwisdom joined #git
17:37 ResidentBiscuit joined #git
17:37 ss22ever22 joined #git
17:39 hahuang65 joined #git
17:40 nettoweb joined #git
17:46 jimi_ joined #git
17:47 sea-gull1 joined #git
17:49 barduck joined #git
17:49 kerrick joined #git
17:49 harish_ joined #git
17:50 MrRoot left #git
17:52 jantje joined #git
17:53 Sasazuka joined #git
17:53 kyledr joined #git
17:54 kerrick joined #git
17:56 kerrick joined #git
17:58 Darren_ joined #git
17:58 djb-irc joined #git
18:01 le_melomane joined #git
18:01 dsantiago joined #git
18:01 ced9000 joined #git
18:02 Alienpruts joined #git
18:03 chrisgeorge joined #git
18:04 multi_io joined #git
18:04 Lipp joined #git
18:05 ced9000 there are many ways to filter the output of git log. is there a way to, instead of filtering out commits that don't match some criterion, display all commits but highlight those that match whatever you're looking for?
18:05 ced9000 since missing commits make the graph view confusing
18:09 glen Vampire0: i tried kde version of svn2git as well. it did not work as i wanted. sorry don't have more details
18:10 mar77i joined #git
18:11 kerrick joined #git
18:11 glen Vampire0: i briefly recall none of them were able to follow up repo history if svn path was moved in svn server.
18:12 glen Vampire0: i.e this kind of moves: /projects/project/{trunk,branches,tags} -> /superprojects/project/{trunk,branches,tags}
18:13 disi joined #git
18:15 JamesBaxter joined #git
18:16 jbitdrop joined #git
18:16 fedepad joined #git
18:16 djb-irc joined #git
18:19 hashpuppy joined #git
18:19 einnjo joined #git
18:19 nettoweb joined #git
18:20 DJ_Spies joined #git
18:20 joki joined #git
18:22 Vampire0 glen, you can easily configure this with the rules files of the KDE svn2git. I transformed much uglier histories with it already, like cvs2svn into trunk/branches/tags, then moved to project1/{trunk/branches/tags}, then next project again to trunk/branches/tags and moved and so on for many subprojects and even worse history crimes
18:23 kyledr joined #git
18:23 Vampire0 glen, was all easy and fast with the right rules file if you know what happened in the history
18:25 fuchstronaut joined #git
18:26 bmahe joined #git
18:26 yottabyte joined #git
18:26 kerrick joined #git
18:26 yottabyte can git commit -a be paired with git commit -m
18:27 yottabyte and is it like not having to do git add -A?
18:28 jaguarmagenta joined #git
18:29 faux___ joined #git
18:29 leksster joined #git
18:29 Kitty- joined #git
18:31 timvisher joined #git
18:32 duderonomy joined #git
18:33 ResidentBiscuit joined #git
18:33 Lipp_ joined #git
18:33 kerrick_ joined #git
18:34 ResidentBiscuit joined #git
18:35 ResidentBiscuit joined #git
18:35 osse yottabyte: git commit -am 'message here'
18:35 User458764 joined #git
18:35 osse commit -a commits the same stuff as  git add -u && git commit  would
18:35 osse not add -A.
18:36 Limes_ joined #git
18:39 wcpan joined #git
18:41 abh joined #git
18:42 ruby_ joined #git
18:42 asdofindia joined #git
18:45 dhanush joined #git
18:45 e14 joined #git
18:47 kerrick_ joined #git
18:47 nidr0x joined #git
18:49 synthroid joined #git
18:51 titankiller joined #git
18:51 Sonderblade can you git log commits in the order in which they landed at the branch and when they were comitted first?
18:52 hwoarang joined #git
18:53 bket left #git
18:53 cdown joined #git
18:54 thiago the order in which they landed is the default order of git log
18:55 peaches joined #git
18:55 peaches joined #git
18:55 jozwior joined #git
18:56 Sonderblade oh but the Date: field appears to be the creation date
18:58 NwS joined #git
18:58 e3 joined #git
18:58 hwoarang joined #git
18:59 thiago the Date: field is the Author date
18:59 thiago if you want to see the date it was actually committed, use --pretty=fuller
19:00 badpeaches joined #git
19:00 Guest24821 left #git
19:01 salparadise so doing a rebase -i on a pretty large set of commits, having problems that I think are related to the size (since I never have issues with 'merge -> rebase -i' with working branches
19:02 salparadise is there another way to squash my commits and doing a new commit message?
19:02 manuels joined #git
19:03 thiago I routinely rebase over 100 comits
19:03 thiago I've never had problems even with 200
19:03 thiago so it's not the size or the number of commits
19:03 shmuli joined #git
19:04 salparadise hmm fair enough, let me keep trying
19:05 sns joined #git
19:05 abh joined #git
19:05 Akimb joined #git
19:06 cdesai joined #git
19:06 nidr0x joined #git
19:06 jokester joined #git
19:08 Tonkers joined #git
19:09 Tonkers Hey all! I have a live site that I want to manage through a git repository. How do I convert the codebase into a git repository without breaking everything???
19:09 harish__ joined #git
19:10 ced9000 Tonkers: go to the root folder of the codebase, initialize git there, add all the files you want to add, and make a commit
19:11 ced9000 it shouldn't break anything
19:12 dhanush joined #git
19:12 perlpilot unless the web server is configured such that it will start serving files in the .git dir ... that might be breakage :)
19:12 Tonkers how do I make sure that doesn't happen? :o
19:13 Tonkers do I need to add a .gitignore beforehand or something perlpilot ?
19:14 brokensyntax joined #git
19:14 perlpilot Tonkers: the problem I described is not a git problem, but a web server problem.  You'd have to configure your web server appropriately (if it really is a problem)
19:14 WaReZ joined #git
19:15 rpg joined #git
19:16 yqt joined #git
19:17 timvisher joined #git
19:17 e3 Hello, is there a good tool to visualize the 3diff output file?
19:17 dreiss joined #git
19:18 thiago kdiff3
19:18 Tonkers ced9000: I init'd in the codebase dir, and it made the empty repo in the parent directory... is that right? git status shows everything untracked in the parent directory...
19:18 kerrick_ joined #git
19:20 e3 thiago: Unfortunately, kdiff3 doesn't spilt up the file into three columns. Do you know a tool which can do this?
19:20 hahuang65 joined #git
19:21 thiago it does split into three columns
19:21 thiago git mergetool calls it in such a way that it shows all three
19:22 Tonkers nvm i'm dumb
19:25 perlpilot Tonkers: as soon as you turn the existing code into a git repo, you might want to clone it somewhere and then work out a deployment process.
19:26 mycelium joined #git
19:27 e3 thiago: mmh. I am a little confused. So, that means mergetool will reformat the file?
19:28 jozwior joined #git
19:29 vuoto joined #git
19:29 Emperor_Earth joined #git
19:31 MillerBoss joined #git
19:31 ForeignBiscuit joined #git
19:31 magellanicloud joined #git
19:32 ForeignBiscuit joined #git
19:32 hwoarang joined #git
19:32 ForeignBiscuit joined #git
19:33 fuchstronaut joined #git
19:33 jumpman joined #git
19:33 Aasdas joined #git
19:33 jnewt joined #git
19:34 ForeignBiscuit joined #git
19:34 jumpman joined #git
19:34 Aasdas left #git
19:34 ForeignBiscuit joined #git
19:35 ForeignBiscuit joined #git
19:35 Tonkers perlpilot: I've init'd it and committed all relevant files. do I just git clone user@server.com:/path/to/my/.git?
19:35 Tonkers or do I have to do more
19:35 WakiMiko joined #git
19:37 eivarv joined #git
19:38 johnmilton joined #git
19:40 Cavallari joined #git
19:41 folkrav joined #git
19:42 thadtheman joined #git
19:44 peaches joined #git
19:44 peaches joined #git
19:45 mar77i joined #git
19:45 thadtheman I'm looking at a magit cheat sheet. It says ( in a magit-status buffer ) type a i to add a gitignore rule globally and type an I to add a gitignore rule locally. can someone clarify the scope of these two?
19:46 Seveas thadtheman: try a magit channel. magit doesn't come up here too often (maybe once a year :)), so I don't know if anyone in here will have an answer for you
19:47 Seveas might be .git/info/exclude (local) versus .gitignore (global). Or .gitignore (local) versus ~/.gitignore (global)
19:48 nwkj86 joined #git
19:48 cdown joined #git
19:48 tgummerer or .gitignore in the repository root, and .gitignore in the directory the file in question is in.  that's only a guess though, I'm not using magit myself
19:49 blackwind_123 joined #git
19:50 EvilPenguin joined #git
19:51 Seveas it's .git/info/exclude versus .gitignore
19:51 eivarv joined #git
19:51 Seveas https://github.com/magit/magit/blo​b/master/lisp/magit-extras.el#L123
19:51 Anja joined #git
19:52 Seveas also, my braces allergy has set in again reading lisp. Anybody got some antihistamine?
19:52 rpg joined #git
19:52 cirocosta joined #git
19:53 timvisher joined #git
19:54 cirocosta hey everyone, does anyone knows whether it's possible to use 'Authorization: Bearer <token>' when cloning a repository using 'git clone https://<token>@git.com/repo'?
19:55 cirocosta I mean, eventually by changing a 'CURL_OPT' or something
19:55 cirocosta (if not configurable directly on git)
19:55 cirocosta thx in advance!
19:55 kyledr joined #git
19:55 Tonkers perlpilot: are git-upload-pack and git-receive-pack additional packages I need to install on my server?
19:55 Seveas cirocosta: see man git-config, search for http.extraHeader
19:55 gitinfo cirocosta: the git-config manpage is available at http://jk.gs/git-config.html
19:56 cirocosta cool! i'll check it out
19:56 cirocosta thanks a bunch!
19:56 folkrav joined #git
19:56 xaa joined #git
19:57 star_prone joined #git
19:58 Seveas git config http.https://git.com.extraHeader "Authorization: Bearer <token>"
19:59 Seveas Tonkers: they come with git
19:59 Tonkers so I probably just don't have access to them in this jailshell then?
20:00 Tonkers ok
20:00 elisa87 left #git
20:00 lizzin joined #git
20:01 aidalgol joined #git
20:02 mar77i joined #git
20:05 Mobutils joined #git
20:08 synthroid joined #git
20:08 rapha joined #git
20:08 rapha Hi!
20:09 rapha Is there a way, in my local clone of a repository, to completely remove all commits and their data (in the .git directory) older than, say, a week or so without that having any effect on current commits or pushes? I.e., I just want the newest part of the repository to save space on my mobile device...
20:10 Seveas No.
20:11 rapha Okay, that was one of the clearest answers I've ever gotten on IRC.
20:11 le_melomane joined #git
20:11 kyledr joined #git
20:11 thadtheman Well what is the difference between adding a pattern to .gitignore and .git/info/exclude
20:12 rapha So if you want to work on a project you don't own with a 10GB history on a small/cheap/old device, then Git is basically discriminating against you? o_O
20:13 Seveas thadtheman: .gitignore gets committed to the repo, so others will see your changes to it when fetching. .git/info/exclude never leaves your machine
20:14 Seveas rapha: you can do a shallow clone, but you cannot turn a full repo into a shallow one.
20:15 Sonderblade you can delete the .git directory. then it becomes "shallow"
20:16 Seveas a bit too shallow :)
20:17 folkrav joined #git
20:17 _Vi joined #git
20:17 yqt joined #git
20:17 rapha Seveas: "shallow clone", that's something to google! Thank you!
20:21 ToBeCloud joined #git
20:23 kyledr joined #git
20:25 danslo_ joined #git
20:25 preaction_ joined #git
20:25 sfix joined #git
20:25 flavius joined #git
20:26 tinanoo1 joined #git
20:29 jaguarmagenta joined #git
20:29 Lor_ joined #git
20:30 harish_ joined #git
20:31 hahuang65 joined #git
20:33 nidr0x joined #git
20:36 duderonomy joined #git
20:36 converge joined #git
20:36 converge joined #git
20:37 kyan_ joined #git
20:37 le_melomane joined #git
20:38 kyan joined #git
20:39 drizztbsd joined #git
20:41 ahr3n joined #git
20:42 nwkj86 joined #git
20:42 FirewalkR joined #git
20:44 AnarchyAo joined #git
20:45 Fabricio20 joined #git
20:45 krabat joined #git
20:46 gruetzkopf joined #git
20:48 cyphase joined #git
20:48 dondata joined #git
20:49 forgot joined #git
20:52 abra0 joined #git
20:52 wriggle joined #git
20:52 cyan__ joined #git
20:53 nidr0x joined #git
20:53 vF3hNGxc47h8 joined #git
20:53 Guest93726 joined #git
20:56 Shentino joined #git
20:56 dondata left #git
20:57 smokemeakipper joined #git
20:57 smokemeakipper left #git
20:58 flopsec joined #git
20:58 phatypus joined #git
20:59 inflames joined #git
20:59 duderonomy joined #git
20:59 nwkj86 joined #git
21:00 axisys joined #git
21:03 basy joined #git
21:03 CussBot_ joined #git
21:06 ryoshu joined #git
21:06 ryoshu left #git
21:08 barajasfab joined #git
21:09 WayToDoor joined #git
21:13 nwkj86 joined #git
21:17 zartu joined #git
21:22 shmuli joined #git
21:22 kyledr joined #git
21:23 rapha left #git
21:27 snthsnth joined #git
21:28 phatypus is it possible to run git log --graph with certain args to that you can limit the results to a specific set of brances?  e.g. git log --oneline --abbrev-commit [only show 'dev' and dev/feature/ticket-abc']
21:28 phatypus *so that
21:31 phos1 joined #git
21:32 m0viefreak git log --oneline --abbrev-commit dev dev/feature/ticket-ab
21:32 phos1 I have setup 4 repositories, Master -> UAT> DEV->Minor Adjustments. When I try to merge from DEV to Minor adjustments (as dev has some updateS) I am getting “Cannot merge unreleated histories”. What does this mean and how do I fix it?
21:34 fuchstronaut joined #git
21:35 prawnzy joined #git
21:36 sea-gull_ joined #git
21:39 cirocosta joined #git
21:40 jantje joined #git
21:41 glen Vampire0: thanks but no. i've spent already too much time customizing current conversion. inventing all to new system, is not worth the effort
21:41 olalonde_ joined #git
21:42 glen Vampire0: and as said earlier, i tried that importer you keep referring to. it had problems, so i took next one
21:42 Kitty- joined #git
21:42 elastix joined #git
21:43 alphor joined #git
21:44 alphor can I use branches to maintain very closely related files? IE dockerfiles that do the same thing with different base OS'
21:45 alex1a joined #git
21:46 star_prone joined #git
21:46 kyledr joined #git
21:47 mkoskar joined #git
21:48 manuels joined #git
21:50 harish__ joined #git
21:51 worstadmin Is it possible to create a submodule from files in different paths of my repo?
21:52 cirocosta joined #git
21:52 worstadmin The folder structure in my project cannot be modified and some of the files in various folders are exclusively worked on by other teams - and they are filling up the git log
21:54 faux____ joined #git
21:56 preaction worstadmin: yes. you can put submodules anywhere in your repo
21:56 phos1 With GIT should you be able to do a merge both ways (A->B and B->A) depending on wht you are doing?
21:56 worstadmin But can I have one submodule that has files that live in various paths in my main repo?
21:56 preaction worstadmin: no. but symlinks are a thing
21:57 preaction phos1: you can, but what would that achieve?
21:57 worstadmin preaction: symlinks don't help me control the git log - I guess I'll just have to separate out the projects entirely
21:58 phos1 preaction: Working on my local repo, push up to the DEV site where all developers merge into, merge in any conflicts. I would then push dev on to UAT, and pull dev pack down to my local repo to keep working on. Is there a better workflow?
22:00 preaction phos1: once you make a release ready for testing, i would make a release branch. bugfixes are made on master and cherry-picked to the release branch if necessary. so the flow is always master -> release and never release -> master
22:00 phos1 So then once I have done my merge on “Dev”, I would make a new branch to pull down to local and work on?
22:01 drizztbsd joined #git
22:06 styx_ joined #git
22:07 aax joined #git
22:09 Vampire0 glen, obviously is has much more serious problems and the problems you had were just wrong usage based. But hey, I'm not saying you should switch over. I just showed you your options and that there is a much better tool with much better results. I don't really care whether you use it or not, it is your pain. ;-)
22:09 faux___ joined #git
22:11 e14 joined #git
22:13 anonymuse joined #git
22:17 shmuli joined #git
22:24 ruby_ joined #git
22:24 Zoxc joined #git
22:27 AnarchyAo joined #git
22:29 jaafar joined #git
22:29 jaguarmagenta joined #git
22:36 Fabricio20 joined #git
22:38 navidr joined #git
22:41 danslo_ joined #git
22:42 mizu_no_oto joined #git
22:48 hahuang65 joined #git
22:50 Anja joined #git
22:55 blackwind_123 joined #git
22:57 led_belly joined #git
22:57 fedepad joined #git
23:00 Skipp_OSX joined #git
23:01 Skipp_OSX is there a way to checkout a branch and rebase it on top of master in one step?
23:01 nilg joined #git
23:03 Skipp_OSX reason I ask is, by checking out an old branch, then rebasing on top of master, many many files get modified, which means I have a longer compile time while the compiler looks at modification times and says, I need to rebuild all this
23:04 jaafar joined #git
23:04 Skipp_OSX but if I could do both steps at once, fewer files would get modified, just those that truly changed between master and branch and compile time would be reduced significantly
23:05 Kitty- joined #git
23:09 safe joined #git
23:10 Aven joined #git
23:10 Aven can I create a new repository using command line??
23:10 harish_ joined #git
23:11 cspack joined #git
23:12 preaction Aven: git init
23:12 Aven ah
23:15 einnjo joined #git
23:16 Sasazuka__ joined #git
23:17 timvisher joined #git
23:22 snowkidind joined #git
23:22 Lipp joined #git
23:24 tgunr joined #git
23:28 Guest9 joined #git
23:31 dsdeiz joined #git
23:31 dsdeiz joined #git
23:32 Lipp joined #git
23:32 Guest9 joined #git
23:34 mizu_no_oto joined #git
23:34 star_prone joined #git
23:35 fuchstronaut joined #git
23:35 fuchstronaut joined #git
23:36 mablae_ joined #git
23:36 phatypus joined #git
23:41 sssilver joined #git
23:42 EvilPenguin joined #git
23:47 nikivi joined #git

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