Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2017-03-19

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

All times shown according to UTC.

Time Nick Message
00:00 Cabanossi joined #git
00:05 peterbecich joined #git
00:09 ResidentBiscuit joined #git
00:11 shgysk8zer0_ joined #git
00:12 mar77i joined #git
00:12 miczac joined #git
00:14 Calinou joined #git
00:26 a3Dman joined #git
00:27 peterbec` joined #git
00:28 masuberu joined #git
00:29 peterbec` joined #git
00:32 m0viefreak joined #git
00:37 SwiftMatt joined #git
00:38 j416 minn: git blame already does that
00:38 j416 minn: man git-blame look at -C and -M
00:38 gitinfo minn: the git-blame manpage is available at http://jk.gs/git-blame.html
00:40 j416 minn: don't forget to include deletions in your metrics :)
00:43 netj joined #git
00:48 peterbec` joined #git
00:53 tmg joined #git
00:55 chipotle joined #git
00:55 chris_wot joined #git
00:58 brent__ joined #git
00:59 JanC joined #git
01:00 Cabanossi joined #git
01:11 brent__ joined #git
01:11 raijin joined #git
01:12 mar77i joined #git
01:23 raijin joined #git
01:23 ozmage joined #git
01:24 dreiss joined #git
01:25 jeffreylevesque joined #git
01:25 ShapeShifter499 joined #git
01:27 minn j416: That's odd. I tried using those options previously, but still found git blame attributing authorship to the copier (e.g., copying a file and calling git blame with -M and -C). I will have to read the man page a little more closely. Thanks for bringing my attention back.
01:27 minn I have not forgotten deletions :)
01:28 jmesmon joined #git
01:31 moei joined #git
01:34 Lyqyd joined #git
01:36 ReptilianBrain joined #git
01:40 raijin joined #git
01:42 mar77i joined #git
01:44 j416 minn: remember that -C can be specified up to three times
01:44 acetakwas joined #git
01:46 aidalgol joined #git
01:48 matoro joined #git
01:52 toothe joined #git
01:53 toothe I made a clone of a repository and put it on github. I made a ton of changes and committed them to master. How woul dI go about mergeing back from upstream after my changes are accepted?
01:54 toothe or rather
01:54 toothe How would I merge new changes of the original repository and yet keep my changes?
01:57 robattil1 joined #git
01:59 minn j416: It seems like I'm getting different behavior when git detects renames versus copies. I will probably just have to dig into the documentation a little more to determine what's going on.
02:01 stevenbenner joined #git
02:01 amdi_ joined #git
02:04 alezandro joined #git
02:04 pwnz0r joined #git
02:04 qpdb joined #git
02:06 svm_invictvs joined #git
02:09 wxg joined #git
02:09 pks joined #git
02:10 pandeiro joined #git
02:12 mar77i joined #git
02:14 j416 minn: I haven't looked into that in detail; good luck with the digging ;) Also, -w might be useful for you.
02:15 Cabanossi joined #git
02:15 atomicb0mb joined #git
02:20 thebinary joined #git
02:20 alezandro joined #git
02:20 d^sh_ joined #git
02:22 ayogi joined #git
02:23 finalbeta joined #git
02:24 Rodya_ joined #git
02:27 finalbeta1 joined #git
02:30 acetakwas joined #git
02:42 mar77i joined #git
02:46 mizu_no_oto joined #git
02:48 ilbot3 joined #git
02:48 Topic for #git is now Welcome to #git, the place for git help. Just git help, apparently. | We are aware of the SHA-1 collision. See: http://jk.gs/t/a4 | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.12.0 | Getting "cannot send to channel"? /msg gitinfo .voice | This is serious business. https://madeitwor.se/git-jokes
02:49 robattila256 joined #git
02:49 pwnz0r joined #git
02:53 ozmage joined #git
02:53 kbs joined #git
02:54 shinnya joined #git
02:55 tristanp joined #git
02:56 fstd_ joined #git
02:56 jeffreylevesque joined #git
02:59 anuxivm left #git
03:04 Orphis joined #git
03:04 matoro joined #git
03:05 ojdo joined #git
03:06 learning joined #git
03:07 armyriad joined #git
03:08 chipotle_ joined #git
03:09 stoopkid joined #git
03:12 mar77i joined #git
03:14 peterbec` joined #git
03:14 peterbecich joined #git
03:17 todd_dsm joined #git
03:17 lagothrix joined #git
03:17 mizu_no_oto joined #git
03:20 Atm0spher1c joined #git
03:21 chachasmooth joined #git
03:24 muhannad__ joined #git
03:25 Lyqyd_ joined #git
03:29 hhee joined #git
03:30 mizu_no_oto joined #git
03:32 rchavik joined #git
03:41 Vampire0 joined #git
03:41 pks joined #git
03:42 mar77i joined #git
03:46 bulletmark joined #git
03:46 relyks joined #git
03:47 relyks is it possible to change the date of a specific commit but leave all the other commit dates intact?
03:47 tax joined #git
03:48 hexagoxel joined #git
03:49 raijin joined #git
03:52 rwp relyks, Yes. Basically a rebase operation. But all later hashes will change.
03:53 preaction so, not without someone noticing, if that's what you're concerned about
03:53 relyks rwp
03:53 relyks couldn't that break the repo though?
03:54 preaction depends on if anyone's pulled it since and after you !rewrite
03:54 gitinfo 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
03:54 rwp Break the repo?  That depends upon your definition.
03:55 Vortex34 joined #git
03:58 relyks thanks everyone
03:58 anth0ny joined #git
04:00 Vampire0_ joined #git
04:04 learning joined #git
04:05 anth0ny joined #git
04:07 safe joined #git
04:08 _Vi joined #git
04:09 Atm0spher1c joined #git
04:12 mar77i joined #git
04:13 Atm0spher1c joined #git
04:19 arescorpio joined #git
04:20 TbobbyZ joined #git
04:21 TbobbyZ joined #git
04:22 TbobbyZ joined #git
04:26 SwiftMatt joined #git
04:27 Sleepy63 joined #git
04:29 chipotle joined #git
04:33 Anja joined #git
04:35 hexagoxel joined #git
04:40 _Vi joined #git
04:42 mar77i joined #git
04:44 afuentes joined #git
04:52 Anja joined #git
04:55 dsdeiz joined #git
04:59 diogenese joined #git
04:59 Cabanossi joined #git
05:03 learning joined #git
05:07 kurokoleung joined #git
05:08 kurokoleung left #git
05:14 muhannad__ joined #git
05:17 pks joined #git
05:21 dave0x6d joined #git
05:22 Derperperd joined #git
05:26 menip joined #git
05:26 mbrumbelow joined #git
05:27 Param joined #git
05:28 Param left #git
05:29 Cabanossi joined #git
05:33 lucido-cl joined #git
05:36 muhannad__ joined #git
05:38 pks joined #git
05:42 relyks left #git
05:42 mar77i joined #git
05:46 hahuang61 joined #git
05:46 ankit01ojha joined #git
05:53 pks joined #git
05:54 peterbec` joined #git
05:54 peterbecich joined #git
05:59 miha- joined #git
06:00 andrew710 joined #git
06:00 webstrand joined #git
06:02 muhannad__ joined #git
06:12 mar77i joined #git
06:17 mishal joined #git
06:21 a_thakur joined #git
06:25 overlord_tm joined #git
06:27 hexagoxel joined #git
06:34 amrits joined #git
06:35 fmeerkoetter joined #git
06:36 muhannad__ joined #git
06:42 mar77i joined #git
06:50 hexagoxel joined #git
06:59 pks joined #git
07:01 cdown joined #git
07:06 tmg joined #git
07:07 MineCoins joined #git
07:09 bgerber joined #git
07:12 mar77i joined #git
07:17 alezandro joined #git
07:18 cdown_ joined #git
07:24 oleo joined #git
07:25 Sound joined #git
07:27 Atm0spher1c joined #git
07:29 Cassiopaya joined #git
07:33 a_thakur joined #git
07:35 bulletmark joined #git
07:40 drodger joined #git
07:42 mar77i joined #git
07:43 sbulage joined #git
07:47 hahuang61 joined #git
07:48 jeffreylevesque_ joined #git
07:49 fatalhalt joined #git
07:50 bariscant joined #git
07:51 bariscant joined #git
07:51 _ng joined #git
07:52 nostrora joined #git
07:56 sdp joined #git
08:00 schleppel joined #git
08:01 kevr i have a bunch of files that have been deleted but the refs still have information about them, thus it downloads unnecessary data
08:01 kevr how can i clean this up?
08:02 Silox| joined #git
08:05 avar !rewrite
08:05 gitinfo 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
08:06 miha- joined #git
08:07 Calinou what's the difference between --force and --force-with-lease?
08:12 kevr avar: okay, that's fine
08:13 mehola joined #git
08:20 MarioBranco joined #git
08:21 _Vi joined #git
08:21 t0by joined #git
08:23 peterbec` joined #git
08:24 peterbecich joined #git
08:25 zxd joined #git
08:25 zxd hi
08:25 zxd how to create new branch
08:25 zxd git branch foo fatal: Not a valid object name: 'bar'.
08:25 a_thakur joined #git
08:34 pks joined #git
08:40 a_thakur joined #git
08:40 GreyXor__ joined #git
08:43 learning joined #git
08:44 NeverDie joined #git
08:44 hexagoxel joined #git
08:53 dsdeiz joined #git
08:56 _ikke_ Looks like your repo is in an inconsistent state
08:56 _ikke_ what does git symbolic-ref HEAD return?
08:57 freimatz joined #git
09:01 Oatmeal joined #git
09:01 dhollin3 joined #git
09:05 Cabanossi joined #git
09:07 roentgen joined #git
09:08 re1 joined #git
09:09 YuGiOhJCJ joined #git
09:10 CtrlC joined #git
09:11 h12o joined #git
09:14 h12o joined #git
09:15 buzzLightBeer joined #git
09:18 hahuang61 joined #git
09:24 Loomaanaatii joined #git
09:27 Loomaanaatii joined #git
09:31 catsup joined #git
09:31 s1scha joined #git
09:32 catsup joined #git
09:39 kexmex joined #git
09:42 davimore joined #git
09:42 vrach joined #git
09:42 shubhamkrm joined #git
09:43 vrach hi there, wonder if anyone can point me to a proper git tutorial? Interactive if possible.
09:45 pks joined #git
09:46 oskarkv joined #git
09:49 tmg joined #git
09:49 n1ce joined #git
09:54 seishun joined #git
09:54 Loomaana_ joined #git
09:55 miczac joined #git
09:59 jstein_ joined #git
09:59 avar is gitrevisions(7) really the most exhaustive documentation for the @{<datespec>} format? E.g. master@{3 hours ago} ?
09:59 [Brain] joined #git
09:59 glebihan joined #git
10:00 bariscant joined #git
10:00 govg joined #git
10:02 _ikke_ avar: It should be
10:03 fedepad joined #git
10:04 nevodka joined #git
10:05 avar wow, that really needs some better docs then
10:05 mar77i_ joined #git
10:06 avar Nothing points out that @{tea time} works: https://github.com/git/git/commit/a8aca418d6
10:10 kjsaihs joined #git
10:13 osk| joined #git
10:13 _Vi joined #git
10:15 anwer joined #git
10:15 cdown_ joined #git
10:17 rokups joined #git
10:24 peterbecich joined #git
10:24 peterbec` joined #git
10:25 Glooomy joined #git
10:27 mquin joined #git
10:31 mquin joined #git
10:33 Chinggis6 joined #git
10:37 aspiers joined #git
10:38 nowhere_man joined #git
10:39 TbobbyZ joined #git
10:41 rudi_s !tutorial
10:41 gitinfo A list of interactive git tutorials: http://try.github.com
10:41 rudi_s vrach: ^
10:41 vrach ahh
10:41 rudi_s But I'd recommend !progit which is a book actually, but teaches all the relevant parts of git IMO very well.
10:41 gitinfo [!book] There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable
10:42 afuentes joined #git
10:45 m0viefreak joined #git
10:47 mbrumbelow joined #git
10:48 tmg !parable
10:48 gitinfo 'The git parable' provides some good reasoning behind git.  http://tom.preston-werner.com/2009/05/19/the-git-parable.html
10:49 King_Hual joined #git
10:49 dersand joined #git
10:50 aviraldg joined #git
10:50 aviraldg joined #git
10:51 j08nY joined #git
10:52 learning joined #git
10:53 Cabanossi joined #git
10:53 aw1 joined #git
10:53 Jackneill joined #git
10:55 jozwior joined #git
10:55 theoceaniscool joined #git
10:56 manuelschneid3r joined #git
10:59 aielima joined #git
11:02 grayjoc joined #git
11:04 ok91 joined #git
11:06 aax joined #git
11:08 mar77i joined #git
11:09 flobin joined #git
11:09 flobin hi everyone
11:10 flobin I have this repo where, no matter what, it keeps showing this one folder as having modified content whenever I check git status
11:10 flobin as in, before committing, after comitting, etc.
11:10 flobin that folder is also a git repo, does that have something to do with it?
11:10 grawity yes
11:10 grawity that sub-repo is probably tracked as a *submodule*
11:11 grawity which means the parent repo only keeps track of the commit ID, not the actual contents
11:11 grawity so if the sub-repo has uncommitted stuff, the parent repo cannot do anything with that
11:12 flobin aha
11:12 flobin so if I change something in the sub-repo it will just keep showing up as modified content?
11:12 grawity until you commit it *inside* the sub-repo, yes
11:12 flobin right, I see
11:12 flobin thanks grawity!
11:13 cbreak joined #git
11:15 vrach guys, is there a difftool one would recommend for using alongside git? (except meld)
11:15 vdv joined #git
11:15 vrach i would not mind paying
11:19 andrew710 joined #git
11:19 hahuang61 joined #git
11:27 mostlybadfly joined #git
11:32 Vampire0 joined #git
11:34 ozmage joined #git
11:35 xaa joined #git
11:38 UTAN_dev_ joined #git
11:39 Cabanossi joined #git
11:40 jameser joined #git
11:42 re1 joined #git
11:44 BSAlb joined #git
11:47 seishun joined #git
11:47 jameser joined #git
11:49 acetakwas joined #git
11:49 txm joined #git
11:50 drodger joined #git
11:52 Zialus joined #git
11:54 cdown joined #git
11:58 doener joined #git
12:02 flobin joined #git
12:04 miha- joined #git
12:04 peterbec` joined #git
12:04 peterbecich joined #git
12:08 mar77i joined #git
12:10 h12o joined #git
12:18 Vampire0_ joined #git
12:20 Lunatrius joined #git
12:21 kittikit joined #git
12:22 eikhorst joined #git
12:25 Murii joined #git
12:26 borkr joined #git
12:31 sbasso joined #git
12:37 joeco joined #git
12:38 robotroll joined #git
12:38 mar77i joined #git
12:38 zxd _ikke_: git symbolic-ref HEAD refs/heads/master
12:38 zxd I just ran git init
12:39 zxd it's a fresh directory
12:39 zxd git branch foo fatal: Not a valid object name: 'master'.
12:39 Cabanossi joined #git
12:39 zxd do I need to commit first for HEAD to be generated
12:39 grawity you need to commit first for refs/heads/master to be generated
12:39 zxd what if I want to name it other than master
12:41 Seveas git checkout -b some-other-name
12:41 Seveas and then commit
12:41 WantFood Is there a nice way to mirror all my github repos (including private ones) automatically to a local git server?
12:41 grawity or just rename master after committing
12:43 alexandre9099 joined #git
12:43 mehola joined #git
12:43 leeN joined #git
12:44 Seveas WantFood: if you don't mind using a third party tool: install my git-spindle and do: git hub mirror WantFood/*
12:44 cdown_ joined #git
12:44 mehola joined #git
12:44 stoopkid joined #git
12:45 WantFood Seveas: Looks good, thanks :)
12:45 mehola joined #git
12:45 muhannad__ joined #git
12:45 mehola joined #git
12:46 h12o joined #git
12:46 mehola joined #git
12:47 mehola joined #git
12:48 mehola joined #git
12:49 mehola joined #git
12:49 dave0x6d joined #git
12:49 mehola joined #git
12:51 Darion joined #git
12:52 Ascathon joined #git
12:55 rchavik joined #git
12:55 rchavik joined #git
13:01 flobin joined #git
13:02 tvw joined #git
13:05 bariscant joined #git
13:06 peterbec` joined #git
13:06 peterbecich joined #git
13:07 hexagoxel joined #git
13:08 jameser joined #git
13:08 mar77i joined #git
13:09 jnavila joined #git
13:09 Cabanossi joined #git
13:10 grayjoc joined #git
13:12 Balliad joined #git
13:14 Chinggis6 joined #git
13:14 h12o joined #git
13:14 King_Hual joined #git
13:18 kacper joined #git
13:19 kacper Hi! Do you know any public service which supports git-annex?
13:20 hahuang61 joined #git
13:21 marvi_ joined #git
13:21 chipotle joined #git
13:21 jameser joined #git
13:22 andrew710 joined #git
13:22 miczac joined #git
13:22 MarioBranco joined #git
13:23 robattila256 joined #git
13:23 learning joined #git
13:26 wootehfoot joined #git
13:27 sarri joined #git
13:27 sarri joined #git
13:29 sbasso joined #git
13:29 XenophonF joined #git
13:29 brent__ joined #git
13:32 gunnaro joined #git
13:32 Gustavo6046 joined #git
13:34 avar Any git hosting site + amazon s3 account
13:37 ski7777 joined #git
13:38 mar77i joined #git
13:39 eroussel joined #git
13:45 avar kacper: ^
13:47 mda1 joined #git
13:50 nd joined #git
13:50 drodger joined #git
13:50 mehola joined #git
13:54 atomicb0mb joined #git
13:56 YuGiOhJCJ joined #git
13:56 wootehfoot joined #git
13:57 Vampire0 joined #git
13:58 _ngz_ngzz joined #git
14:05 thebird joined #git
14:06 peterbec` joined #git
14:07 peterbecich joined #git
14:07 bariscant joined #git
14:08 zefferno joined #git
14:08 mar77i joined #git
14:16 ankit01ojha joined #git
14:17 CEnnis91 joined #git
14:22 _ngz_ngzz joined #git
14:22 dendazen joined #git
14:23 NullableTruth joined #git
14:31 flobin joined #git
14:32 Vampire0_ joined #git
14:37 nwmcsween_ joined #git
14:38 mar77i joined #git
14:39 Ardethian\ joined #git
14:40 sixtel left #git
14:41 Kaisyu joined #git
14:41 grayjoc joined #git
14:44 zxd hi
14:44 zxd I changed the author name in ~/.gitconfig then did git commit  --amend -m 'msg'
14:44 zxd but still it uses the old author name
14:45 King_Hual joined #git
14:45 grawity that's normal; --amend only updates committer name
14:46 zxd how to update author name
14:47 zxd I only see  authoer any way with git log
14:47 zxd how to see the commiter name
14:48 avar --reset
14:48 dviola joined #git
14:51 madewokherd joined #git
14:52 a_thakur joined #git
14:52 nwmcsween__ joined #git
14:58 publio joined #git
15:00 Balliad joined #git
15:00 lucido-cl joined #git
15:01 dreiss joined #git
15:03 _ng joined #git
15:05 _ikke_ --reset-author
15:06 eahmedshendy joined #git
15:06 eahmedshendy joined #git
15:08 shinnya joined #git
15:08 mar77i joined #git
15:09 amdi_ joined #git
15:11 eclecticjohny joined #git
15:20 Greench joined #git
15:20 hahuang61 joined #git
15:22 Greench hello here! I have just pushed a commit to my branch, and only now I see that my branch is 3 commits behind master. how should I proceed to include these 3 commits in my branch without losing my changes?
15:23 j416 Greench: rebase or merge; pick one
15:23 Cabanossi joined #git
15:24 flobin joined #git
15:24 j416 Greench: I typically rebase feature branches on top of whatever they are based off of to avoid merge conflicts.
15:25 j416 Greench: (and by that I mean, merge commits with resolved merge conflicts, sorry for wording that oddly)
15:27 Greench thanks j416, I tried git rebase origin mybranch , it says mybranch is uptodate
15:27 Greench am I doing it wrong?
15:29 j416 Greench: man git-rebase
15:29 gitinfo Greench: the git-rebase manpage is available at http://jk.gs/git-rebase.html
15:30 j416 scroll down to the first example
15:30 j416 you want: git rebase master
15:31 j416 assuming your situation is what that graph depicts
15:33 Isla_de_Muerte joined #git
15:33 Greench git rebase master and git rebase origin give the same output, mybranch is up to date
15:33 j416 origin is not a branch
15:34 j416 don't confuse remotes with branches
15:34 j416 if you don't have the latest local copy of master, you can do: git rebase origin/master
15:34 Greench oh, okay, I'll try to remember that
15:34 j416 or, you can git checkout master && git pull && git checkout - && git rebase master
15:34 theoceaniscool joined #git
15:34 j416 to get up to date with the latest master first and then rebase atop that
15:35 muhannad__ joined #git
15:35 Yewbacca joined #git
15:36 nostrora joined #git
15:36 Greench I see. It was comparing to my local master not the remote master
15:37 waveclaw joined #git
15:37 Yewbacca Hi, I'm trying to be a good developer. I want my commit summary and my message to be appropriate. So first of all, what's right for summary? strlen(commitsummary) < 50 or <= 50? I always stop at 49 because I don't know if 50th column is ok.
15:37 t0by joined #git
15:37 t0by joined #git
15:38 zxd if committer and author are the same then git log won't show committer name?
15:39 j416 git log will only show author name unless you specify a different format
15:41 shine1 joined #git
15:41 grayjoc joined #git
15:42 zxd ok
15:42 deltab Yewbacca: the git-commit man page says "Though not required, it’s a good idea to begin the commit message with a single short (less than 50 character) line summarizing the change, followed by a blank line and then a more thorough description."
15:42 ayogi joined #git
15:43 zxd what's the acceptable way for setting repositories where they are all under a common subject or theme  how to group them together?
15:43 deltab Yewbacca: so, less than 50, though it's not a hard limit
15:43 Yewbacca deltab: Thanks. "Less than 50 character", yeah... that suggests < 50 (so 1-49 characters). But I just checked Github and it uses "<= 50". And tpope (one of the people who helped set the rule?) says <= 50.
15:43 Yewbacca http://tbaggery.com/2008/04/19/a-note-about-git-commit-messages.html
15:43 * deltab nods
15:44 deltab I didn't know it said "less than" until I looked
15:44 Yewbacca What do you think I should stick with? 50 I guess. 49 seems kinda odd.
15:44 deltab yeah, 50
15:44 _ikke_ 50 is not a hard limit
15:44 Yewbacca I know but it's something to do with "git log" in an 80 character terminal, not breaking lines
15:44 Cthalupa joined #git
15:44 grayjoc joined #git
15:44 zxd is it possible to  do like   git clone   parent-folder/project1    , git clone parent-folder/project2
15:44 zxd I guess so why not
15:44 Yewbacca Thanks deltab
15:44 zxd silly me
15:45 Yewbacca The other thing I worry about is the line width for the actual message? <= 72 is the number that keeps coming up. Sounds right or wrong?
15:48 Yewbacca ls
15:48 Yewbacca Am I still connected?
15:48 j416 TIL "git rebase" is valid as-is without extra arguments. Has it always been like that? o_O
15:48 deltab that's from an email convention that allows three levels of quoting with "> " within 80 columns
15:49 j416 handy.
15:49 vuoto joined #git
15:49 Yewbacca deltab: Hi again, I got disconnected after this:
15:49 Yewbacca The other thing I worry about is the line width for the actual message? <= 72 is the number that keeps coming up. Sounds right or wrong?
15:49 Yewbacca Anything after that is not in my chat log :P
15:50 deltab 15:48:32 < Yewbacca> ls
15:50 deltab 15:48:37 < Yewbacca> Am I still connected?
15:50 deltab 15:48:49 < j416> TIL "git rebase" is valid as-is without extra arguments. Has it always been like that? o_O
15:50 deltab 15:48:49 < deltab> that's from an email convention that allows three levels of quoting with "> " within 80 columns
15:51 Yewbacca Ahh thanks. And it's right? <= 72?
15:51 deltab j416: no it hasn't
15:52 deltab Yewbacca: yes (but again, just convention)
15:52 Yewbacca deltab: Thanks a lot, I'll set up my editor now. It's been running in freeflow mode and I've been using the text cursor to check how many characters I had. It was hell xD
15:52 deltab make it about the same width as the template text
15:52 Yewbacca I noticed I can set automatic limits for commits, so it's time to do that.
15:53 Yewbacca Ah yes smart
15:55 deltab vim has syntax highlighting that applies colour near the end of the line, rather than a hard limit
15:55 Yewbacca Yeah I use vim
15:55 Yewbacca But switched to Spacemacs (vim in Emacs)
15:55 d0nn1e joined #git
15:55 Ascathon joined #git
15:55 Yewbacca And Magit. Lovely module.
15:56 j416 deltab: thanks
15:57 j416 deltab: do you happen to know since when?
16:00 Sound joined #git
16:00 Cthalupa joined #git
16:00 deltab sometime after 1.7.4
16:01 vuoto joined #git
16:02 j416 huh, pretty early on still
16:02 deltab here: https://github.com/git/git/commit/15a147e61898d25ec8b539190e87f3a09592c9c8
16:02 j416 thanks for digging!
16:04 j416 not sure I've ever had use for this particular thing though come to think of it
16:04 j416 good to know it's there at least.
16:05 deltab that commit message makes a case for it
16:05 robotroll joined #git
16:06 j416 yes, that's what I'm referring to
16:06 Greench j416: so, I have done as you said, I switched to local master, pulled the new stuff, then rebased. After a bit of fiddling, I managed to push my branch. Now it says I'm 3 commits behind and 3 commits ahead of master. Is that normal? :/
16:06 j416 Greench: read again
16:07 j416 Greench: or, what did you do?
16:07 Greench the fiddling part? :-$
16:07 j416 Greench: if you did "git checkout mybranch && git rebase master" then you should be N commits ahead of master
16:07 peterbecich joined #git
16:07 peterbec` joined #git
16:08 j416 Greench: it sounds like you did "git checkout master && git rebase mybranch"
16:08 mar77i joined #git
16:09 Greench when I run your command again, it says lots of stuff then "no changes, patch already applied".
16:09 j416 what command?
16:09 Greench "git checkout mybranch && git rebase master"
16:10 j416 I'd advise you to stop for a moment and examine the state of things.
16:10 j416 did you run what I guessed above?
16:10 Greench perhaps at some point, let me check...
16:11 j416 anyway,
16:11 j416 first !backup
16:11 gitinfo Worried about your data while trying stuff out in your repo? The repository in its entirety lives inside the .git directory in the root of your work tree so to backup everything `cp -a path/to/workdir path/to/backup` or equivalent will suffice as long as the repo is not modified during backup. See also http://sethrobertson.github.com/GitBestPractices/#backups
16:11 wootehfoot joined #git
16:11 j416 then use caution and: git fetch && git checkout master && git reset --hard @{u} && git checkout mybranch && git rebase master
16:12 j416 Greench: if my understanding of your problem is correct, the above should do it. If not, restore from the backup you made.
16:14 sbasso joined #git
16:14 t0by joined #git
16:14 eahmedshendy joined #git
16:17 Greench j416: so I backed up as you said, I ran the command, and when I go to the gitlab page, it still shows 3 | 3
16:17 j416 Greench: you didn't push it yet
16:17 j416 Greench: check the log locally, then force-push it
16:17 miha- joined #git
16:18 j416 Greench: or, remember the commit has that is in gitlab so that you can get back to that easily; then "git checkout mybranch && git push -f origin mybranch"
16:19 seishun joined #git
16:21 Greench j416: when you say "remember the commit", is that related to a git command? :-$
16:22 j416 Greench: it means look at the commit hash and memorise it. Write it down or something.
16:22 Greench ah, that I can do
16:22 j416 make a song about it so that it'll become easier to recall.
16:23 j416 etc
16:23 j416 as long as you have the hash of a commit and that commit exists locally, you can access it.
16:23 kbs joined #git
16:24 Greench ah, okay. I fail hard at understanding git I'm afraid... I need to find me a flowchart or something..
16:24 j416 did you try !book ?
16:24 gitinfo There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable
16:25 j416 the basic concept is very easy.
16:25 j416 once you grasp that, it's only a matter of learning what commands alter the state of the repo in the way you want.
16:25 flobin joined #git
16:25 a3Dman joined #git
16:26 Greench thanks :)
16:27 Greench 0 commits behind, 1 commit ahead, yay
16:27 j416 correct?
16:27 nothingnew joined #git
16:28 kbs out of just curiosity, I've been poking at the 'patience/histogram' diff algorithm in (c)git and jgit. To my eyes anyway, it seems like it basically wants to do a weighted subsequence match, rather than a longest subsequence match. Why isn't it just... doing this - I feel I must be missing something :)
16:29 Greench all good, muchas gracias j416!
16:29 j416 Greench: cool
16:29 j416 hth
16:29 jozwior joined #git
16:29 Murii joined #git
16:32 Gsham joined #git
16:35 chuchunaku joined #git
16:42 learning joined #git
16:43 Greench left #git
16:45 svm_invictvs joined #git
16:46 diogenese joined #git
16:47 coco joined #git
16:51 a_thakur joined #git
16:51 a_thakur joined #git
16:52 eikhorst joined #git
16:52 pickfire joined #git
16:53 mehola joined #git
16:55 Lyqyd joined #git
16:57 thebird joined #git
16:58 JanC joined #git
16:58 Wind0r joined #git
17:02 dreiss joined #git
17:05 XVar joined #git
17:08 peterbec` joined #git
17:08 peterbecich joined #git
17:08 mar77i joined #git
17:08 a3Dman joined #git
17:09 roentgen joined #git
17:10 Orphis joined #git
17:10 dviola joined #git
17:12 pickfire Hi, I am Ivan Tham. I am planning to apply for gsoc 2017 for git.
17:12 multi_io joined #git
17:13 flobin left #git
17:15 JPisaBrony joined #git
17:16 venmx joined #git
17:16 Gsham joined #git
17:17 stfn joined #git
17:20 bariscant joined #git
17:20 joki joined #git
17:21 IRCFrEAK joined #git
17:21 hahuang61 joined #git
17:23 a_thakur joined #git
17:25 menip joined #git
17:25 ankit01ojha joined #git
17:29 miczac joined #git
17:30 mknod joined #git
17:36 seishun joined #git
17:38 d1z joined #git
17:38 bariscant joined #git
17:38 mar77i joined #git
17:39 _ikke_ Hello
17:40 Glooomy joined #git
17:40 TbobbyZ joined #git
17:43 ameivq joined #git
17:43 _ng joined #git
17:46 hasc joined #git
17:48 Doginal joined #git
17:52 YuGiOhJCJ joined #git
17:53 eihabarabia joined #git
17:56 drodger joined #git
17:57 alezandro joined #git
18:03 svm_invictvs joined #git
18:04 Skyfinn joined #git
18:05 SpeakerToMeat joined #git
18:08 mar77i joined #git
18:10 d^sh joined #git
18:11 SpeakerToMeat joined #git
18:11 ertesx joined #git
18:11 raijin joined #git
18:13 chipotle joined #git
18:17 Skyfinn joined #git
18:23 pks joined #git
18:23 Skyfinn joined #git
18:24 Cabanossi joined #git
18:24 lucido-cl joined #git
18:28 venmx joined #git
18:30 sbasso joined #git
18:33 alezandro joined #git
18:36 roentgen joined #git
18:38 mar77i joined #git
18:39 cdown_ joined #git
18:41 Skyfinn joined #git
18:42 brent__ joined #git
18:43 drodger joined #git
18:45 venmx joined #git
18:48 nowhere_man joined #git
18:50 Es0teric joined #git
18:53 nowhere_man joined #git
18:53 paul424 joined #git
18:54 mehola joined #git
18:54 dreiss joined #git
18:54 drodger joined #git
18:56 dvaske joined #git
18:59 sbasso joined #git
18:59 j416 o/ Ivan
19:02 Shakeel_ joined #git
19:04 envian joined #git
19:04 byte512 joined #git
19:05 cr34ton joined #git
19:06 learning joined #git
19:07 Atm0spher1c joined #git
19:09 mar77i joined #git
19:09 drodger joined #git
19:09 Balliad joined #git
19:09 mizu_no_oto joined #git
19:10 peterbec` joined #git
19:10 warthog9 joined #git
19:10 peterbecich joined #git
19:11 kexmex joined #git
19:18 cdown joined #git
19:18 aidalgol joined #git
19:19 Glooomy joined #git
19:19 Ardethian joined #git
19:22 hahuang61 joined #git
19:23 a_thakur joined #git
19:24 Vinnie_win_s joined #git
19:26 warthog9 joined #git
19:28 Atm0spher1c joined #git
19:30 anuxivm joined #git
19:30 dirtyroshi joined #git
19:31 nwmcsween__ joined #git
19:34 safe joined #git
19:37 nowhereman joined #git
19:38 mar77i joined #git
19:43 wootehfoot joined #git
19:43 BlackMaria_netsp joined #git
19:44 _ng joined #git
19:45 acetakwas joined #git
19:46 a3Dman joined #git
19:48 dviola joined #git
19:49 mellernoia joined #git
19:50 zerow joined #git
19:52 no_gravity joined #git
19:53 cagmz joined #git
19:54 drodger joined #git
19:54 Cabanossi joined #git
19:54 TbobbyZ joined #git
19:57 acetakwas joined #git
20:00 no_gravity I wonder why the command to switch to a branch is called "checkout".
20:01 regedit joined #git
20:01 _ikke_ It checks out files from the repository
20:02 Vampire0 no_gravity, why not, how else would you call the action "switch current branch and update working directory files to the branch version"?
20:02 nwmcsween_ joined #git
20:03 no_gravity Is "checkout" a term that was used for similar actions before git?
20:03 dstolfa no_gravity: Not at all.
20:03 dstolfa no_gravity: checkout traditionally meant what git calls clone, more or less
20:03 no_gravity Then my first thought how to name the command would have been "switch" I guess.
20:04 dstolfa no_gravity: git commands are god awful and unintuitive, but you get used to it, and git itself is pretty good
20:04 _ikke_ no_gravity: It does more than just switch branches
20:04 bremner dstolfa: I think that's just a bit subjective
20:04 dstolfa bremner: Might be, but a large number of people, myself included, consider the git cli a disaster, despite git itself being good(which is why I use it for all my development)
20:05 dstolfa The cli itself by no means diminishes the technology, you can have any number of frontends for it
20:05 _ikke_ dstolfa: The only way to solve it is to severely limit the cli
20:05 dstolfa _ikke_: Simply renaming the commands to something more intuitive would go a long way.
20:06 _ikke_ dstolfa: Then you would get an explosion of commands, and still having problems finding good names
20:06 Vampire0 dstolfa, that is hiiighly subjective. I myself like the CLI and find the commands pretty intuitive mostly
20:07 no_gravity What is the idea behind "checkout"? Like "Check it out"?
20:07 no_gravity As in "Try it out" or "Take a look at it"?
20:07 TbobbyZ joined #git
20:07 dstolfa Vampire0: Well, all I can say is good for you, I've not taught git to a single person that didn't comment on the names of git cli commands and how god awful they are
20:08 Vampire0 no_gravity, checkout was used by other VCS before already to update the working tree to a specific state
20:08 dstolfa Regardless whether that person is coming from 0 knowledge of a VCS, whether they are coming from svn/cvs and what not
20:08 dstolfa That's not to say that some other VCS are any better
20:08 no_gravity Vampire0: So it refers to that?
20:08 dstolfa hg isn't much better in terms of that
20:08 jnavila joined #git
20:09 mar77i joined #git
20:09 Vampire0 no_gravity, I don't know, I didn't invent it, I just say that SVN and other VCS named a very similar action the same before
20:09 jamick joined #git
20:09 Goplat joined #git
20:09 demize no_gravity: Think of it in the same way as checking something out from a library.
20:10 _ikke_ checkout files from the repository in the working tree
20:10 no_gravity demize: I'm not sure what it means to check out something from a library. Does it mean lending something?
20:10 Vampire0 dstolfa, I'm not gonna discuss with you, I just told you my opinion. There are many people loving the Git CLI and I am one of them. That you only tried to teach Git to people that have a different meaning could be coincidence, it could be caused by how you teach them, or it could be caused by anything else, I don't really care. ;-)
20:10 dstolfa Vampire0: svn checkout and git checkout don't really have much correspondence to eachother though
20:11 dstolfa svn checkout is going to check out a working copy from a repository, git checkout can create branches.
20:11 _ikke_ dstolfa: they both checkout files
20:11 _ikke_ from the repository in your working tree
20:11 learning joined #git
20:12 no_gravity I feel like it's time to become fluid with git because I am building a distributed team. And git is the version control most coders are using these days.
20:13 dstolfa no_gravity: You should be using a DVCS either way, there are a select few cases where you wouldn't want a DVCS, but those are very few. git is a good one to go with
20:14 _ikke_ !books
20:14 gitinfo [!book] There are several good books available about git; 'Pro Git' is probably the best: http://git-scm.com/book but also look at !bottomup !cs !gcs !designers !gitt !vcbe and !parable
20:14 no_gravity dstolfa: I developed my own vcs 10 years ago and it works wonderfully til this day.
20:14 no_gravity dstolfa: But it's faster to onboard new coders when you use tools they already know.
20:15 dstolfa no_gravity: Well, svn works wonderfully as well, the question is what you expect from a VCS to provide to you, do you want file integrity, do you want a working repository on every developer's machine, do you want ease of branching and so on. Right tool for the job.
20:15 _ikke_ for some values of "working" :P
20:15 jozwior joined #git
20:16 Vampire0 dstolfa, that "git checkout" can create branches is just an additional convenience shortcut. This has nothing to do with the meaning of the command
20:16 dstolfa _ikke_: Well that's true of every software out there.
20:16 _ikke_ hehe
20:16 dvaske joined #git
20:17 _ikke_ dstolfa: I used svn at my previous company, but it was a mess imo
20:17 no_gravity I'm a bit confused about how merge conflicts look in the log.
20:17 no_gravity So when you resolve a merge conflict and commit the result, there seems to be no changepatch in that commit.
20:17 dstolfa _ikke_: I still use svn in some cases, they're mostly cases where you would need to prove that a revision you killed does no exist in any repository any more
20:17 dstolfa But I agree, it's a mess for the most part
20:18 dstolfa Especially when it comes to branching
20:18 _ikke_ dstolfa: How about people still having that revision checked out?
20:19 dstolfa _ikke_: They have it checked out, but not the actual repository
20:19 dstolfa They might clone the repository and mirror it, for example
20:19 _ikke_ right, but they could commit it back in right away
20:19 dstolfa But I don't think that falls under the same legal categories
20:20 dstolfa _ikke_: They could, it ultimately comes down to legalities more than anything
20:20 _ikke_ alright
20:20 dstolfa And I'm not really able to discuss the differences here because I don't know why one is fine and other isn't
20:20 dstolfa But that's why some companies and projects opt for non-DVCS
20:21 _ikke_ Another reason is when you work with non-mergable files, requiring locks etc
20:21 dstolfa (not that I particularly agree with that, but *shrug*)
20:21 dstolfa _ikke_: Not being able to commit things locally and then push it all together is a PITA on it's own
20:21 _ikke_ yup
20:21 _ikke_ then you get these monster commits
20:22 manuelschneid3r joined #git
20:22 trewe joined #git
20:22 dstolfa _ikke_: Well, I often see monster commits through git as well, but that's more of a case where people don't know how to use a DVCS
20:23 _ikke_ right
20:23 _ikke_ But in git, you don't have any excuses ;-)
20:23 nowhereman joined #git
20:23 dstolfa _ikke_: That's probably why they go for svn, or better yet, cvs :7
20:24 _ikke_ haha
20:24 dstolfa Or well, if you're absolutely insane, RCS
20:24 sythe joined #git
20:25 jnavila joined #git
20:28 xaviergmail joined #git
20:28 mehola joined #git
20:29 raijin joined #git
20:29 jagob joined #git
20:31 seishun joined #git
20:32 no_gravity I created a new repo. When I do "git branch disaster" I get "fatal: Not a valid object name: 'master'." - what does that mean?
20:33 anth0ny joined #git
20:34 _ikke_ no_gravity: when you create a new repository, there do not exist any branches yet.
20:34 btree joined #git
20:34 no_gravity _ikke_: Oh, I thought "git branch disaster" would creat a branch called "disaster"?
20:35 _ikke_ no_gravity: branches always point to commits, but because there is no commit yet, the branch is not yet 'valid'
20:35 nowhereman joined #git
20:35 no_gravity You cannot branch before the first commit?
20:36 Cxcf joined #git
20:36 btree left #git
20:36 Cxcf Hey all
20:37 MarioBranco joined #git
20:37 kadoban no_gravity: Things work a bit weird before the first commit, it's kind of an edge case.
20:38 no_gravity kadoban: I see. Thanks.
20:38 kadoban You can always just create an empty "initial" commit.   'git commit --allow-empty -m "initial commit"'
20:38 _ikke_ kadoban: thanks for explainig it, I was looking for a good way to describe it
20:38 no_gravity Then how can every be repo be equal?
20:38 no_gravity If one has the initial commit?
20:38 no_gravity Isn't that the godfather repo then?
20:38 kadoban Hmm?
20:38 _ikke_ no_gravity: Not following you
20:38 no_gravity Linus promised all repos are equal.
20:39 mar77i_ joined #git
20:39 no_gravity But now I learn that one has to make the initial commit.
20:39 kadoban I think you're misunderstanding something. I'm not sure what though.
20:40 kadoban Have you read any beginner materials? Some experience just using things will likely help make things clearer
20:40 no_gravity kadoban: In this case I think you misunderstand something.
20:41 _ikke_ no_gravity: history has to start somewhere
20:41 no_gravity It's a philosophical question.
20:41 no_gravity _ikke_: Isn't the owner of the first commit god then?
20:41 kadoban It doesn't sound like a particularly meaningful one.
20:41 _ikke_ no_gravity: No, why would he?
20:42 no_gravity _ikke_: "The creator".
20:42 no_gravity The artificer of the universe.
20:42 no_gravity Who turns *nothing* into *something*.
20:42 _ikke_ anyone can create a root commit
20:42 _ikke_ a repository can even contain multiple root commits
20:43 kadoban git is a piece of software designed for a fairly specific task. If you're looking for religion, git is unlikely to help.
20:43 no_gravity _ikke_: I see. So maybe only the master branch is a special branch.
20:43 _ikke_ Not really, it's just that when you just start a repository, it points to nothing
20:43 _ikke_ no_gravity: git commit knows that edge case, creates a commit without a parent
20:44 _ikke_ but when git branch tries to create a new branch, it does not know what to point it to
20:45 _ikke_ git branch tries to dereference master, but master is not valid yet
20:45 _ikke_ git branch checks HEAD, HEAD points to master, master points to nothing
20:45 no_gravity Strange, I added a line in "master", did not commit it, did "git checkout disaster" and the line is in the disaster branch too.
20:46 anth0ny joined #git
20:46 _ikke_ uncomitted changes don't belong to any branch
20:46 Discovery joined #git
20:46 _ikke_ they exist in your working tree
20:46 _ikke_ !float
20:46 gitinfo If you have made a change in your working directory and have NOT YET COMMITTED, you may "float" that change over to another (`git checkout oldbranch`) or new (`git checkout -b newbranch`) branch and commit it there.  If the files you changed differ between branches, the checkout will fail.  In that case, `git stash` then checkout, and `git stash apply` and go through normal conflict resolution.
20:47 peterbec` joined #git
20:47 no_gravity Wuups! That changes my whole perspective of "branches".
20:47 peterbecich joined #git
20:47 no_gravity I have to digest that. Not sure what to think of it...
20:47 _ikke_ you commit on branches, those changes stay there
20:47 _ikke_ anyhting not comitted yet, is not part of history
20:48 dstolfa no_gravity: The concept is largerly the same, except the changes are local, and branches point to actual commits
20:49 dstolfa For example, unstaged changes, i.e., non-committed changes, don't belong to the repository(simplified), so you could change the branch at will, and commit them there, causing that branch to point to your new commit
20:49 no_gravity Strange.. what is the use case for that?
20:50 dstolfa no_gravity: To me it looks more of a side-effect as to how branches work, as opposed to going out of one's way to implement that
20:50 _ikke_ git checkout just looks at what files are different from the previous commit, and updates those
20:51 no_gravity If I work on a Game and have 2 branches "Zombie killer" and "Zombie tamer" .. wouldn't I want to be able to switch back and forth between them and try them out? Witout the obligation to have everything commited all the time?
20:51 thebird joined #git
20:51 dstolfa no_gravity: Well, no, because commit in git doesn't automatically mean push or rebase, as it does with svn
20:52 kbs joined #git
20:52 no_gravity What I mean is: The way it works, I *cannot* switch between the "killer" and the "tamer" version unless I have all changes commited.
20:52 _ikke_ no_gravity: Note that commits in git are cheap
20:52 _ikke_ no_gravity: also checkout man git worktree
20:52 gitinfo no_gravity: the git-worktree manpage is available at http://jk.gs/git-worktree.html
20:53 thebird joined #git
20:53 no_gravity So if I am dabbling with the headshot sound in the killer version - which is a headpad sound in the tamer version - I *cannot* switch to the tamer version or it will have the wrong sound.
20:53 no_gravity Unless I commit the sounde.
20:54 dstolfa no_gravity: Pretty much
20:54 dstolfa You could stash, for example
20:54 no_gravity But what if I want to switch back and forth because I don't know if I want to commit the sound?
20:54 dstolfa man git stash
20:54 gitinfo the git-stash manpage is available at http://jk.gs/git-stash.html
20:54 Cabanossi joined #git
20:54 Cxcf My understanding of a good workflow for personal projects: (1.) Create/fork repo (2.) create a dev branch (3.) create your feature branch (4.) change, stage, commit (5.) git pull from dev to make sure/correct conflicts (6.) once dev branch is how you like it, pull from master (7.) fix any conflicts (8.) push/pull request into master.  Is there anything wrong or unusual about my workflow?
20:54 venmx joined #git
20:56 _ikke_ Sounds good, but it will result in a lot of merge commits
20:56 dstolfa _ikke_: You could rebase
20:56 _ikke_ right
20:56 _ikke_ But not everyone is comfortable with rebasing
20:56 dstolfa Well, they should be :P
20:56 a3Dman joined #git
20:57 _ikke_ I agree
20:57 no_gravity Are merge commits only neccessary when there are merge conflicts or will they be created for every merge?
20:57 Cxcf _ikke_: what steps should I elimate to decrease merge conflicts?
20:57 dstolfa no_gravity: Every merge
20:58 _ikke_ no_gravity: everytime history diverged
20:58 mda1 joined #git
20:58 no_gravity That sounds ugly.
20:58 dstolfa no_gravity: It is, until you find out about git-rebase
20:58 _ikke_ no_gravity: it allows git to know what has been merged
20:58 canton7 they're created for every non-fast-forward merge
20:58 WayToDoor joined #git
20:58 canton7 git will default to fast-forward merging, if it can
20:59 no_gravity Damn, I started with git an hour ago and already I need git init, git add, git commit, git log, git branch, git stash, git merge and git rebase...
20:59 no_gravity If it continues at that pace, I will have nothing but git commands in my head by the end of next week.
20:59 dstolfa no_gravity: That was kind of the point I was making earlier, but *shrug*
20:59 _ikke_ no_gravity: And you will probably need some more too
20:59 no_gravity dstolfa: What point?
20:59 _ikke_ no_gravity: But that's what allows git's flexibility
20:59 dstolfa no_gravity: That the cli is a mess
21:00 _ikke_ dstolfa: So you're complaining when you need to use more then a screw driver and a hammer for your building project?
21:00 canton7 make sure you get to the point where you visualise what each command does to the tree of commits
21:00 canton7 everything makes when you "get" that
21:00 _ikke_ +sense
21:00 canton7 oops, I've been travelling for the last 8 hours ><
21:00 dstolfa _ikke_: There's the git GUI which is far more friendly than the cli, it _can_ be done with the cli, as I often have my own scripts that reuse the git cli
21:00 raijin joined #git
21:00 dstolfa Another example is magit, it works really well.
21:00 no_gravity dstolfa: It seems it's the git concept that makes all that neccessary.
21:01 _ikke_ dstolfa: I always get anoyed with GUIs because the always limit what you can do
21:01 roentgen joined #git
21:01 _ikke_ They often use their own (imo confusing) terminology
21:01 dstolfa _ikke_: Check out magit, it lets you do everything you need to, it's far cleaner than the git cli
21:01 no_gravity I wonder if it is really neccessary that every merge results in a merge commit.
21:02 no_gravity Well, probably yes.
21:02 SCHAAP137 joined #git
21:02 _ikke_ no_gravity: that's how git maintaines a sane DAG
21:02 no_gravity _ikke_: DAG?
21:02 canton7 every merge *doesn't* result in a merge commit
21:02 canton7 only non-fast-forward merges
21:02 dstolfa no_gravity: directed acyclic graph
21:02 no_gravity canton7: What's a fast-forward-merge?
21:02 canton7 read !bottomup or !cs for a graph-cased description of git
21:02 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)
21:02 gitinfo "Git for Computer Scientists" is a quick introduction to git internals for people who are not scared by phrases like Directed Acyclic Graph. http://eagain.net/articles/git-for-computer-scientists/ See also !concepts !bottomup
21:02 _ikke_ !ff
21:02 gitinfo A fast-forward merge occurs when you merge a commit which is a descendant of !HEAD. No new commit is created, instead the branch is simply moved forward. See http://sandofsky.com/images/fast_forward.pdf
21:02 canton7 no_gravity, !ff
21:02 canton7 bleh, what was that trigger...
21:02 _ikke_ canton7: you had the correct one, I was quicker :P
21:02 canton7 aah, of couse
21:03 canton7 *course
21:03 no_gravity Is there a good audio book about git?
21:03 _ikke_ I'm not aware of one
21:04 _ikke_ no_gravity: Those merge commits let's git know what commits have been merges, an what commits haven't, allowing you to keep on merging branches
21:04 _ikke_ have been merged*
21:04 no_gravity "allowing you to keep on merging branches"?
21:05 _ikke_ You can continue comitting on a branch after it has been merged
21:06 irc3k joined #git
21:06 no_gravity I don't see how that is related to merge commits?
21:06 _ikke_ no_gravity: A merge commit is a commit with 2 parents, 'tying' both branches together
21:06 no_gravity Sure
21:08 _ikke_ when you merge 2 diverged branches together, you _have_ to create a commit, containing the merge result
21:08 _ikke_ whether there were conflicts or not has very little to do with that
21:08 no_gravity Ok.
21:09 _ikke_ You can create that commit on only one branch (which git calls a squash commit), but you loose the information about that branch being merged in
21:09 no_gravity When you do "git log" is it possible to see which commits were done in which repo?
21:09 _ikke_ No, a commit does not belong to a repo
21:09 _ikke_ A repository has no identity
21:09 no_gravity Well, it was *done* in a repo.
21:10 no_gravity So if I have a repo "zombie_killer" and a repo "zombie_tamer" and I merge commits from killer to tamer, the info gets lost that the commits were done in killer?
21:10 MineCoins joined #git
21:11 gunnaro_ joined #git
21:11 raijin joined #git
21:11 canton7 yeah. history is just a series of commits. It doesn't matter what repo they were committed in, merged in, etc
21:11 _ikke_ no_gravity: git records what _branch_ it came from
21:11 no_gravity Uh. I don't like that.
21:11 no_gravity Well, in this case it's not branches. It's repos.
21:11 _ikke_ a repo is basically a branch
21:12 _ikke_ If you make a commit on master in one repository, you diverged from the other repositories
21:13 _ikke_ from master in the other repositories
21:13 jost__ joined #git
21:13 no_gravity But that info gets lost?
21:14 canton7 it was never recorded
21:14 canton7 repositories don't have an "identity"
21:15 jstimm joined #git
21:16 jnavila joined #git
21:16 dsdeiz joined #git
21:16 dsdeiz joined #git
21:17 dviola joined #git
21:17 _ikke_ You're basically building a datastructure in a decentralized matter
21:17 no_gravity I'm confused. So I develop a nice aiming-module in repo zombie_killer. And then I pull that into repo zomebie_tamer. Then the info that the aming-module was developed in zombie_killer is lost?
21:18 canton7 it was never recorded
21:18 _ikke_ no_gravity: You record that it was created in a *branch* called zombie_tamer
21:18 _ikke_ that's what matters
21:18 no_gravity _ikke_: I don't have such a branch.
21:18 canton7 you don't pull branches into repos
21:18 no_gravity I only have 2 repos. Both only have a master branch.
21:19 dstolfa no_gravity: Is there a reason why you would want to work with different repositories as opposed to branches?
21:19 _ikke_ no_gravity: You can still create a zombie_killer branch in that other repo
21:19 no_gravity dstolfa: Both repos are playable games. Reachable at zombiekiller.nogravity.com and zombietamer.nogravity.com
21:19 dstolfa no_gravity: Ah, okay, makes sense.
21:19 canton7 no_gravity, you should almost certainly be working with different branches, rather than different repositories. I think your mental model of branches, repos, and how they fit together is lacking. That's fine, but it's worth working on that before trying to ask the sorts of questions you're asking: they'll answer themselves as your understanding improves
21:20 _ikke_ no_gravity: What is the relationship between those to repos then? Why merge at all?
21:20 canton7 if they have shared history - which they have to, in order to merge - then they're the *same* project. The "repos" distinction is just two people working on the same project, effectively
21:20 _ikke_ You can also merge non-related repositories together
21:21 ferr1 joined #git
21:21 _ikke_ git doesn't mind, it just does not have the information to properly merge changes together
21:21 canton7 I thought it refused? Must be misremembering
21:21 _ikke_ git does not care at all
21:21 saidi joined #git
21:22 no_gravity _ikke_: Well, it's the same game. Developed by two developers.
21:22 bariscant joined #git
21:22 svm_invictvs joined #git
21:22 _ikke_ no_gravity: You just said there were 2 games
21:23 no_gravity _ikke_: In the sense that each game can be played at any time.
21:23 hahuang61 joined #git
21:24 thebird joined #git
21:24 Cabanossi joined #git
21:24 _ikke_ no_gravity: Note that even though both repos have a branch called master, they're basically just 2 branches that you need to merge
21:24 a_thakur joined #git
21:24 _ikke_ no_gravity: also read about !workflow
21:24 gitinfo no_gravity: Finding the right workflow for you is critical for the success of any SCM project.  Git is very flexible with respect to workflow.  See http://sethrobertson.github.com/GitBestPractices/#workflow for a list of references about choosing branching and distributed workflows.
21:26 no_gravity _ikke_: I would assume the workflow is like this: the coder of the "killer" version brings up a new feature. After everybody played both versions for a while, everybody is convinces the feature rocks. So the coder of the "tamer" version pulls it via "git pull /game/versions/killer/".
21:26 _ikke_ no_gravity: in git, killer would be a branch
21:26 trewe joined #git
21:26 _ikke_ which you can merge into 'mainline'
21:26 _ikke_ whatever mainline is for you
21:27 noahsussman left #git
21:27 no_gravity _ikke_: But why would the repo of the "tamer" coder need to have a "killer" branch? What if he didn't even know the "killer" coder has been coding on a version of the game?
21:27 raijin joined #git
21:28 _ikke_ no_gravity: He would only get that branch when he fetches it from somewhere
21:29 Cxcf hey guys, what are all the !<keword> things you keep mentioning?
21:29 no_gravity _ikke_: Let's say its in /coders/killer/myversion/.
21:29 _ikke_ Cxcf: !trigger
21:29 gitinfo Cxcf: [!triggers] Please don't spam me! I'm just a poor bot! Here's everything I know: http://jk.gs/git/bot/trigger.php -- that's a list of these nifty keywords like "!bot" that you can use in the channel, in case you were wondering.
21:29 canton7 presumably it would be a branch for the new feature, rather than the whole of killer?
21:29 no_gravity _ikke_: And the killer coder simply coded away on the master branch that he cloned from /coders/tamer/myversion/
21:29 no_gravity _ikke_: Now how does tamer coder go about pulling?
21:29 xaviergmail joined #git
21:30 _ikke_ no_gravity: Note that even though he works on master, it's basically still a branch he works on
21:30 no_gravity _ikke_: So how does tamer pull?
21:30 _ikke_ just not an aptly named branch
21:31 _ikke_ no_gravity: What many people use is a shared repository where you can push to and fetch from
21:31 no_gravity _ikke_: How would that solve the situation of killer and tamer?
21:32 _ikke_ no_gravity: killer would push his branch to the shared repo, tamer can fetch from that repo, and choose to merge it or not
21:33 _ikke_ no_gravity: by explicitly using branches (and not all working on master), this will be a lot more clear
21:33 davimore_ joined #git
21:33 no_gravity _ikke_: But to use branches, we don't need the shared repo. Why did you introduce it?
21:33 _ikke_ no_gravity: I assume both tamer and killer start from some shared state? (git would also assume this)
21:34 sbasso_ joined #git
21:34 _ikke_ no_gravity: The shared repo was just to be able to share branches with eachother asynchornously. It's not required to create and work on branches
21:34 jeffreylevesque joined #git
21:34 _ikke_ asynchronously*
21:36 _ikke_ no_gravity: Just to give you some context, a typical workflow in git would be that a user clones his repo from somehere, starts working on it (on a separete branch or not)
21:36 girassolbit joined #git
21:36 _ikke_ Then depending on the project, he can request someone to pull from him, or push to a shared repo
21:36 _ikke_ so all uses start from the same base repo
21:37 no_gravity _ikke_: I will play around with that.
21:37 no_gravity Thanks for the info.
21:39 learning joined #git
21:39 anth0ny joined #git
21:41 charlenopires joined #git
21:42 raijin joined #git
21:43 masuberu joined #git
21:44 nd joined #git
21:45 no_gravity _ikke_: In the workflow you propose, could the killer coder stay on the killer branch all the time or would there be a need to checkout to other branches at times?
21:45 Droolio joined #git
21:46 Es0teric joined #git
21:47 nd joined #git
21:49 d1z joined #git
21:50 dvaske joined #git
21:50 _ikke_ no_gravity: git is not really made to perpatually keep things branched (while also incorporating new changes from other branches)
21:51 _ikke_ You check out other branches when you want to use it, work on it, or merge into it / rebase it
21:54 oskarkv joined #git
21:54 Cabanossi joined #git
21:54 dreiss joined #git
21:57 nowhereman joined #git
21:57 alexandre9099 joined #git
21:58 charlenopires joined #git
22:02 raynold joined #git
22:03 orbyt_ joined #git
22:05 sbasso joined #git
22:08 no_gravity _ikke_: Now I'm completely confused :)
22:09 netj joined #git
22:09 drodger joined #git
22:09 nanga joined #git
22:14 no_gravity Interest in version control drastically drops during the holidays: https://trends.google.com/trends/explore?q=svn,git,github
22:15 TbobbyZ joined #git
22:18 tmg joined #git
22:20 jamick joined #git
22:21 Gsham joined #git
22:26 acetakwas joined #git
22:27 hahuang61 joined #git
22:29 a3Dman joined #git
22:30 netj joined #git
22:32 Es0teric joined #git
22:34 nowhereman joined #git
22:35 Gsham joined #git
22:38 askb joined #git
22:40 hahuang61 joined #git
22:43 bitmod joined #git
22:45 menip joined #git
22:46 anth0ny joined #git
22:46 Es0teric joined #git
22:50 dvaske joined #git
22:50 lucido-cl joined #git
22:54 dingbat left #git
22:55 dingbat joined #git
22:55 dingbat left #git
22:55 sbasso joined #git
22:56 overlord_tm joined #git
22:59 kexmex joined #git
23:01 Es0teric joined #git
23:01 livingstn joined #git
23:02 anth0ny joined #git
23:07 watabou joined #git
23:11 a3Dman joined #git
23:12 Es0teric joined #git
23:14 acetakwas joined #git
23:19 Naan joined #git
23:19 raijin joined #git
23:22 drodger joined #git
23:23 re1 joined #git
23:25 a_thakur joined #git
23:25 sword joined #git
23:27 Kaisyu joined #git
23:27 navidr joined #git
23:32 hasc left #git
23:32 acetakwas joined #git
23:36 Gsham joined #git
23:37 cbreak joined #git
23:41 Teny joined #git
23:42 a3Dman joined #git
23:43 acetakwas joined #git
23:44 Sound joined #git
23:45 roelmonnens joined #git
23:46 irqq joined #git
23:47 BlackPanx joined #git
23:47 planigan joined #git
23:48 prg3 joined #git
23:48 sespiros joined #git
23:49 overlord_tm joined #git
23:53 Cabanossi joined #git
23:56 peterbecich joined #git
23:57 bitmod joined #git
23:59 xaviergmail joined #git

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