Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-11-24

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

All times shown according to UTC.

Time Nick Message
00:00 nettoweb joined #git
00:00 kyan joined #git
00:00 synx Hey, in a merge commit is there anyway to tell which files had conflicts and had to be resolved? Does git maintain that information?
00:01 dsantiag_ joined #git
00:01 osse synx: I think the default message contains their names
00:01 osse try 'git show mergecommit' does it show a diff?
00:01 multi_io osse: hm
00:02 osse multi_io: but usually the dst part in 'git fetch remote src:dst' is not a branch, but rather something like refs/remotes/blah/bleh.
00:03 osse aka a remote tracking branch
00:03 osse aka the stuff you want local branches to track
00:03 multi_io osse: git branch -u lbranch remote/rbranch   ==> fatal: branch 'remote/rbranch' does not exist
00:03 multi_io ok...
00:03 osse try refs/remotes/remote/rbranch
00:04 cdown joined #git
00:05 multi_io fatal: Couldn't find remote ref refs/remotes/remote/rbranch
00:06 multi_io which kinda makes sense... it doesn't exist locally when the command starts
00:06 Olipro joined #git
00:06 multi_io ah, wait
00:06 multi_io I tried git fetch, not git branch
00:06 osse multi_io: as an example, master usually tracks origin/master. IOW: refs/heads/master usually tracks refs/remotes/origin/master. but refs/remotes/origin/master physically speaking is local thing. fetch's usual job is to update that thing to match the actually remote branch. So it makes sense for fetch not not care much about tracking info since it supplies what's to be tracked in the first place
00:08 osse !remote_tr
00:08 gitinfo [!remote_tracking_branch] Remote-tracking branches (branches which start with e.g. 'origin/', listed by 'git branch -r') are read-only mirrors of the corresponding branches in another repository. They're updated by 'git fetch'. You can't edit them directly (trying to check them out results in a !detached HEAD), but you can create a new local branch based on a remote-tracking branch using e.g. 'git checkout -b <branch> <remote
00:09 GavinMagnus left #git
00:09 dreiss joined #git
00:09 synx osse thanks for the help, in an annoying twist of fate my server just went down so I'll have to give it a shot later, but I appreciate the pointer!
00:11 jfelchner joined #git
00:11 nettoweb joined #git
00:11 multi_io yeah, got it now
00:12 multi_io git branch -u refs/remotes/pub-origin/master pub-master
00:12 multi_io worked. thanks osse
00:13 dendazen joined #git
00:13 osse oh, i got the args the other way around
00:14 osse sorry
00:14 kaiver_ joined #git
00:15 bathtub_shark joined #git
00:16 nettoweb1 joined #git
00:16 Corosauce joined #git
00:16 t-mart_ joined #git
00:16 peepsalot joined #git
00:16 mischat joined #git
00:16 GavinMagnus1 joined #git
00:17 mischat joined #git
00:17 los2 joined #git
00:18 mischat joined #git
00:18 meloc joined #git
00:19 meloc joined #git
00:19 mischat joined #git
00:19 d5sx43 joined #git
00:19 mischat joined #git
00:20 mischat joined #git
00:20 d0nn1e joined #git
00:21 multi_io synx: that's not a twist of fate, that's Murphy's law :P
00:21 mischat joined #git
00:22 osse Murphy's appendix: If something can go wrong but *doesn't* then time will tell that it would have been better if it went wrong in the first place
00:22 kaiver_ osse: :'D
00:23 kadoban osse: Nice
00:24 osse synx: anyway, my point is that 'git show mergecommit' usually doesn't show a diff. But if it does I *think* that's the conflicting parts. Or maybe you need --cc or something like that
00:24 Electrometro joined #git
00:25 GavinMagnus joined #git
00:26 GavinMagnus1 joined #git
00:26 imachuchu joined #git
00:27 nettoweb joined #git
00:28 imachuchu is there an easy way to generate patches from git to apply on a machine with the patch command (it doesn't have git installed)? I've attempted with git format patch, but those patches are for git's consumption and not really patches
00:28 dsantiago joined #git
00:29 kadoban imachuchu: git diff makes acceptable patches I believe
00:29 osse it does
00:29 venmx joined #git
00:29 Siegfried joined #git
00:29 d^sh joined #git
00:30 umpc joined #git
00:30 osse but you can edit the format-patch generated ones. cut off everything before the first 'diff --git ...' and the last two lines.
00:32 synx osse ok, I'll start from there and read the manual for it. Again, appreciate it.
00:33 synx and multi_io yea luckily it's a fantastic excuse to stop working early :)
00:33 GavinMagnus joined #git
00:33 dreiss joined #git
00:33 r00twrh joined #git
00:34 GavinMagnus joined #git
00:36 GavinMagnus left #git
00:36 done joined #git
00:37 Electrometro joined #git
00:38 GavinMagnus1 joined #git
00:39 multi_io when I have lbranch tracking remote/rbranch, does this give me any convenience shortcuts if lbranch isn't checked out?
00:41 multi_io e.g. if I want to push the latest lbranch to remote/rbranch, it seems I still have to git push remote lbranch:rbranch
00:42 osse multi_io: git push remote lbranch should work
00:42 multi_io just as if I didn't have the tracking relationship
00:42 Darren_ joined #git
00:42 multi_io osse: no, that creates a new remote/lbranch
00:43 texinwien_ joined #git
00:43 multi_io (and leaves remote/rbranch where it is)
00:44 GavinMagnus1 left #git
00:44 osse then i don't know
00:44 cqi joined #git
00:45 multi_io ok, thanks
00:45 multi_io maybe it's just convenience for the checked out branch then
00:46 GavinMagnus joined #git
00:46 Siegfried joined #git
00:47 bathtub_shark joined #git
00:47 osse multi_io: what does ''git config branch.lbranch.remote'' say?
00:47 bathtub_shark joined #git
00:48 cdown joined #git
00:48 GavinMagnus left #git
00:48 Rodya_ joined #git
00:50 imachuchu ok, so you can apply patches created with format-patch just fine. All you need to do is when applying the patch use the "-p1" option to patch (assuming that the system being deployed to has the same relative locations as the patch generating system). Thanks everyone for your help!
00:51 kuldeep joined #git
00:51 imachuchu you also don't even need to remove anything from the patch, patch will eat it just fine (besides the mentioned flag for slashes)
00:52 osse oh right
00:53 osse patch is usually good at ignoring noise in the patch file
00:53 rockstar__ joined #git
00:56 multi_io osse: remote
00:57 kuldeep Hello, i have a project which uses modified upstream code (using submodule). i usually update to latest by rebasing my branch. but since rebase and force updating make the previous branch history vanish, so someone checkout at older commit in my will not be able to get the changes of submodule since they were forced updated. i have though of releasing tag xyz-1, xyz-2, .... xyz-N annotated on every rebase of my rebase branch and in
00:57 multi_io $ git config branch.pub-master.remote
00:57 kuldeep .gitsubmodules file: use branch = tag/xyz-N to use that instead of a branch. good idea?
00:57 multi_io pub-origin
00:58 multi_io (in my real case, lbranch = master, rbranch = pub-master, remote = pub-origin)
00:59 kuldeep *older commit in my code (other than HEAD) will
00:59 robattila256 joined #git
01:00 Siegfried joined #git
01:02 osse multi_io: what does ''git config branch.lbranch.merge'' say?
01:02 sarbs joined #git
01:02 osse (hmm, I think this is just what git branch -vv will tell you)
01:02 kuldeep basically, i want to keep all commit of the submodule (modified upstream) that i used in the main project. if i dont, checking out older version of code will have no crossponding commit in submodule since they are rebased and force updated.
01:03 jumpman joined #git
01:04 madewokherd joined #git
01:04 osse kuldeep: Not entirely sure what you mean, but git submodule update will make sure that the submodule has the right commit checked out. You should do that after merge, pulls and rebases
01:05 osse 9Oh I see...
01:06 osse Tagging sounds fine.
01:06 osse Even better if you can get your changes submitted upstream :P
01:06 osse Brb sleep
01:06 kuldeep osse, most of them are specific to my project
01:08 dreaddy joined #git
01:08 kuldeep do i need to update .gitsubmodule file with: branch = tag/xyz-N     everytime i release xyz-N tag.   or i can just use branch = xyz  (xyz branch will contain the latest code that will be released in with next value N+1)
01:10 goTAN joined #git
01:10 osse I don't know. I think I'd do it the old fashioned way, by committing the change to the submodule.
01:10 Skryll joined #git
01:10 synchronous joined #git
01:10 osse The tag is still useful. It keeps the old commits referenced
01:11 warthog9 joined #git
01:12 pikajude joined #git
01:12 kuldeep but im not sure if i have to update .gitsubmodule everytime (with branch = tag/xyz-N) when i release a tag.
01:12 Electrometro joined #git
01:13 kuldeep example: since the branch xyz will not contain commit of xyz-1  (branch xyz will contain the latest changes). how will git resolve the commit-id?
01:23 mischat joined #git
01:24 kuldeep thanks osse
01:26 bathtub_shark joined #git
01:29 oskarkv joined #git
01:31 Rodya_ joined #git
01:36 manfredn joined #git
01:38 ResidentBiscuit joined #git
01:39 legumbre left #git
01:48 Kaisyu joined #git
01:51 Pie_Napple joined #git
01:54 Electrometro joined #git
01:56 eblip joined #git
01:58 Shentino joined #git
02:00 cdown joined #git
02:00 a_thakur joined #git
02:08 finalbeta joined #git
02:08 bathtub_shark joined #git
02:09 Shentino joined #git
02:11 fubu joined #git
02:12 AfroThundr joined #git
02:20 XHFHX joined #git
02:20 d^sh joined #git
02:20 XHFHX Hi there. Can someone help me to remove a commit? I know that you shouldn't remove a commit, but I made it with the wrong email adress and now I comitted with a personal account where i don't want to commit with a personal account
02:21 XHFHX I used bitbucket
02:21 waterCreature joined #git
02:23 Gamecubic joined #git
02:28 GavinMagnus joined #git
02:28 XHFHX anyone?
02:29 GavinMagnus left #git
02:30 diegoaguilar joined #git
02:31 XHFHX removed the branch, not a big problem since it was the first commit on it anyway
02:31 pks joined #git
02:32 ayjay_t joined #git
02:32 dodobrain joined #git
02:33 Rodya_ joined #git
02:33 refried_ joined #git
02:34 diegoaguilar I want to make a commit which will be immediately rebased by fixup to previous commit
02:34 diegoaguilar and I achieve that in a single git command?
02:36 ResidentBiscuit joined #git
02:38 davidfetter_ge diegoaguilar, if i understand correctly, you make the new commit, then rebase from two commits back and squash the second
02:39 diegoaguilar yep, I usually do that
02:39 diegoaguilar but ... Id like to know if theres a more direct way
02:39 davidfetter_ge i'm pretty sure i'd want that to have some way to inspect what's going on
02:39 Dougie187 left #git
02:39 davidfetter_ge even if i did it quickly
02:39 davidfetter_ge !squash
02:40 davidfetter_ge hrm. i guess that's not a keyword
02:41 snowkidind joined #git
02:42 rewt do you mean make the new commit as if it was part of the previous commit?
02:46 inflames joined #git
02:46 railssmith joined #git
02:46 fubu joined #git
02:46 venmx joined #git
02:48 ilbot3 joined #git
02:48 Topic for #git is now Welcome to #git, the place for git help and serious business | Public logs at http://jk.gs/glog | First visit? Read: http://jk.gs/git | Current stable version: 2.10.2 | Getting "cannot send to channel"? /msg gitinfo .voice | The git-jokes project has been suspended for lack of contribution. Help revive it! https://madeitwor.se/git-jokes
02:51 steelsw0rd joined #git
02:52 bufu joined #git
02:53 diegoaguilar joined #git
02:55 ochorocho__ joined #git
02:55 bannakaffalatta joined #git
02:59 goTAN joined #git
03:00 dsantiago joined #git
03:04 stnutt joined #git
03:04 Sazhen86 joined #git
03:04 Nilesh_ joined #git
03:05 aavrug joined #git
03:06 sunri5e joined #git
03:07 shgysk8zer0 joined #git
03:11 ResidentBiscuit joined #git
03:12 ResidentBiscuit joined #git
03:12 chachasmooth joined #git
03:14 cyan__ joined #git
03:15 oaaao joined #git
03:17 hexagoxel joined #git
03:18 bathtub_shark joined #git
03:23 _UniFreak joined #git
03:25 mischat joined #git
03:25 CrazyBonz joined #git
03:29 harrymoreno joined #git
03:30 Andrew_K joined #git
03:33 howdoi joined #git
03:33 Sasazuka joined #git
03:34 Rodya_ joined #git
03:34 jaguarmagenta joined #git
03:35 A5101 joined #git
03:35 Pseudocrat joined #git
03:35 Pseudocrat joined #git
03:38 danslo joined #git
03:40 Goplat joined #git
03:40 BackEndCoder joined #git
03:44 lb1c joined #git
03:47 crater2150 joined #git
03:47 AfroThundr joined #git
03:48 Vampire0 joined #git
03:51 chamunks joined #git
03:53 jerrcs joined #git
03:53 fubu joined #git
03:55 phanimahesh joined #git
03:55 Peetz0r joined #git
03:57 ExeciN joined #git
03:57 Lyqyd joined #git
03:59 lindenle joined #git
04:00 lindenle joined #git
04:00 jimi_ joined #git
04:01 a_thakur joined #git
04:02 fubu joined #git
04:04 sunri5e joined #git
04:10 ResidentBiscuit joined #git
04:14 Vortex35 joined #git
04:15 venmx joined #git
04:17 shgysk8zer0 joined #git
04:21 infernix joined #git
04:22 markvandenborre joined #git
04:22 clemf joined #git
04:22 stnutt left #git
04:25 jaguarmagenta joined #git
04:25 ResidentBiscuit joined #git
04:26 gfixler joined #git
04:26 mischat joined #git
04:27 markvandenborre I have a git repo with a pile of json files in it. Those are basicly job files for batch processing videos. The idea is that geeks attending our FOSS conference basicly fill out the missing information in those json files and commit.
04:28 romerocesar joined #git
04:28 Vampire0_ joined #git
04:30 markvandenborre I wondered about the cleanest way to detect which of the json files have changed. While I was writing this, I was wondering if the solution would be in git, but just by writing this out, now I know that inotify-tools would probably be a cleaner solution.... :-)
04:30 markvandenborre so thanks, even without replies, you have been very useful to me, people in the #git channel :-)
04:30 markvandenborre left #git
04:31 nowhereman joined #git
04:31 fubu joined #git
04:35 Rodya_ joined #git
04:40 valkyr2e joined #git
04:45 rwb joined #git
04:47 hhee joined #git
04:48 goTAN joined #git
04:48 donmichelangelo joined #git
04:48 fstd joined #git
04:49 subhojit777 joined #git
04:50 ResidentBiscuit joined #git
04:50 ayogi joined #git
04:57 ochorocho__ joined #git
04:58 ResidentBiscuit joined #git
05:00 buzzLightBeer joined #git
05:04 bufu joined #git
05:07 rscata joined #git
05:08 d0nn1e joined #git
05:12 Cabanossi joined #git
05:19 NeverDie joined #git
05:25 Peng_ joined #git
05:26 mischat joined #git
05:27 justanotheruser joined #git
05:32 bocaneri joined #git
05:33 justanotheruser joined #git
05:35 thebope_ joined #git
05:35 cagedwisdom joined #git
05:35 pigeon joined #git
05:36 Rodya_ joined #git
05:36 cagedwisdom joined #git
05:36 mizu_no_oto joined #git
05:40 acln joined #git
05:41 jimi_ joined #git
05:41 venmx joined #git
05:43 cdg joined #git
05:48 NeverDie_ joined #git
06:05 a_thakur joined #git
06:09 cagedwisdom joined #git
06:13 LOMAS joined #git
06:17 lindenle joined #git
06:22 sdothum joined #git
06:26 ngperfex joined #git
06:32 rendar joined #git
06:34 LukasT joined #git
06:35 JrCs joined #git
06:35 qt-x joined #git
06:36 LukasT Hi, is there some way to check if git has some text files checked in with CRLF (instead of LF)? On checkout they are CRLF, but I need to verify git object store/git index [I'm not sure why term is correct or if it is two different things]
06:36 moei joined #git
06:37 Rodya_ joined #git
06:37 goTAN joined #git
06:38 milki LukasT: no theres no built-in tool to do that
06:39 meauses joined #git
06:40 snthsnth joined #git
06:41 venmx joined #git
06:42 LucaTM joined #git
06:44 cyan__ joined #git
06:46 finalbeta joined #git
06:46 overlord_tm joined #git
06:48 hahuang65 joined #git
06:48 ThomasLocke joined #git
06:50 aidalgol joined #git
06:52 Darren_ joined #git
06:53 acln joined #git
06:54 clemf joined #git
06:56 jaguarmagenta joined #git
06:56 ome joined #git
06:56 LukasT Is it possible to turn off line normalization? autocrlf true will checkout files as they are checked in?
06:58 dar123 joined #git
06:59 _ikke_ you want autocrlf false
07:00 Rodya_ joined #git
07:01 Raging_Hog joined #git
07:02 cyphase joined #git
07:02 venmx joined #git
07:03 Jameser joined #git
07:07 snthsnth joined #git
07:08 vuoto joined #git
07:08 zeroed joined #git
07:11 rominronin joined #git
07:15 JeroenT joined #git
07:16 adac joined #git
07:19 cagedwisdom joined #git
07:22 dendazen joined #git
07:24 nidr0x joined #git
07:24 meauses joined #git
07:25 ronskisika joined #git
07:25 barduck joined #git
07:26 Guest2 joined #git
07:27 chele joined #git
07:29 mischat joined #git
07:31 Siegfried joined #git
07:32 Alienpruts joined #git
07:36 freimatz joined #git
07:38 Sceorem joined #git
07:41 done joined #git
07:42 jaguarmagenta joined #git
07:44 govg joined #git
07:45 overlord_tm joined #git
07:46 hhee joined #git
07:46 xall joined #git
07:46 rominronin joined #git
07:50 robattila256 joined #git
07:50 JJjack joined #git
07:53 snthsnth joined #git
07:53 JeroenT joined #git
07:54 a_thakur joined #git
07:56 thebope_ joined #git
07:58 nedzadhrnjica joined #git
07:59 danslo joined #git
08:02 jnavila joined #git
08:02 Rodya_ joined #git
08:03 venmx joined #git
08:03 JeroenT joined #git
08:04 freimatz Hi, is there a way in sourcetree to look back to older logs? i.E i pushed something and then i see the commands and output from git commands. can i get back these text after closing this window?
08:06 rominronin joined #git
08:06 dminuoso joined #git
08:06 dminuoso freimatz: Well you have the reflog
08:06 adac joined #git
08:08 milki joined #git
08:09 freimatz there are not all messages i had in sourcetree
08:11 _ikke_ I don't think source tree stores those messages
08:14 TomyWork joined #git
08:16 MrMojito joined #git
08:16 danslo joined #git
08:16 breezy_ joined #git
08:17 breezy_ If i downloaded somebody's repository with a history of all their commits, and i decide to checkout one of their tags, and alter-save some stuff... is there anyway to get back to the original repo?
08:18 mischat joined #git
08:19 lindenle joined #git
08:19 snthsnth joined #git
08:22 a_thakur joined #git
08:22 freimatz _ikke_: i think too, but where are them
08:22 brk_ joined #git
08:23 Anja joined #git
08:25 PCatinean joined #git
08:29 acetakwas joined #git
08:29 CrazyBonz joined #git
08:30 ProbabilityMoon joined #git
08:32 ytacious joined #git
08:34 Yotta92 joined #git
08:34 Macaveli joined #git
08:36 goTAN joined #git
08:36 ivh_ left #git
08:38 texinwien_ joined #git
08:38 _ikke_ freimatz: You have to ask source tree
08:38 _ikke_ It's a proprietary tool, we know very little about it :-)
08:40 cagedwisdom joined #git
08:41 elect joined #git
08:41 snthsnth joined #git
08:42 InfoTest joined #git
08:42 grayjoc joined #git
08:43 byte512 joined #git
08:44 SwiftMatt joined #git
08:46 Snugglebash joined #git
08:46 encod3 joined #git
08:46 morfin joined #git
08:46 morfin hello
08:46 inflames joined #git
08:47 morfin is there simple way to look at history of some lines in Git?
08:47 osse git log -L
08:48 nedzadhrnjica joined #git
08:49 dminuoso breezy_: Sure.. just "git checkout branch" and you're back
08:49 morfin ah i remember now
08:50 freimatz i often use blame
08:50 breezy_ dminuoso: so if i git checkout 2a, and then i open some files and modify and save, then want to see the original 2a, i do "git checkout branch 2a"?
08:51 dminuoso breezy_: no, you just "git stash"
08:51 breezy_ ok thx'
08:51 breezy_ probably have to read the git book lol
08:51 dminuoso breezy_: And if you want your work back you can just pop/apply the stashed changes.
08:52 dminuoso breezy_: Yes, probably.
08:52 dminuoso breezy_: Or better yet branch off and commit those changes and use branches.
08:52 dminuoso I usually only use stash when im in the middle and can't commit yet for reasons.
08:52 breezy_ ok ya i will have to learn how to do all that
08:53 breezy_ right now i only know how to checkout and commit/push
08:54 mikecmpbll joined #git
09:00 ochorocho__ joined #git
09:02 zulutango joined #git
09:02 mindriot101 joined #git
09:03 Rodya_ joined #git
09:03 ahmedelgabri joined #git
09:03 venmx joined #git
09:04 halbomin joined #git
09:07 fees joined #git
09:08 rendar joined #git
09:09 kurkale6ka joined #git
09:11 barduck joined #git
09:12 djb-irc joined #git
09:12 JeroenT joined #git
09:14 hhee joined #git
09:15 Sceorem joined #git
09:18 marcogmonteiro joined #git
09:18 berndj joined #git
09:19 encod3_ joined #git
09:21 zeroed joined #git
09:21 zeroed joined #git
09:24 root3 joined #git
09:24 Sceorem joined #git
09:24 encod3 joined #git
09:24 irco joined #git
09:25 chll_ joined #git
09:27 skunkz joined #git
09:28 Sceorem_ joined #git
09:29 venmx joined #git
09:30 jknetl joined #git
09:30 a_thakur joined #git
09:32 ahmed_elgabri joined #git
09:32 cagedwisdom joined #git
09:38 rominronin joined #git
09:39 clmsy joined #git
09:40 JeroenT joined #git
09:40 root3 left #git
09:42 Serpent7776 joined #git
09:42 mischat joined #git
09:44 JeroenT joined #git
09:44 brk_ joined #git
09:46 ocbtec joined #git
09:51 zeroed joined #git
09:51 zeroed joined #git
09:55 acetakwas joined #git
10:00 davimore joined #git
10:01 JJjack joined #git
10:01 JeroenT joined #git
10:02 Qria joined #git
10:02 cagedwisdom joined #git
10:02 JeroenT joined #git
10:02 InfoTest joined #git
10:03 JeroenT joined #git
10:03 done joined #git
10:04 cagedwisdom joined #git
10:04 Rodya_ joined #git
10:04 seishun joined #git
10:06 dodobrain joined #git
10:06 govg joined #git
10:06 cagedwisdom joined #git
10:07 mablae joined #git
10:11 danslo_ joined #git
10:11 rnsanchez joined #git
10:20 rominronin joined #git
10:20 CrazyBonz joined #git
10:23 jaguarmagenta joined #git
10:24 bathtub_shark joined #git
10:24 miko joined #git
10:25 goTAN joined #git
10:25 cdown joined #git
10:28 CrazyBonz joined #git
10:29 dsdeiz joined #git
10:31 SwiftMatt joined #git
10:32 mischat joined #git
10:33 thebope_ joined #git
10:35 go1111111 joined #git
10:37 CrazyBonz joined #git
10:37 ugjka joined #git
10:37 ferr joined #git
10:41 manuel joined #git
10:41 nettoweb joined #git
10:42 ngrin joined #git
10:44 courrier joined #git
10:45 crazycoder joined #git
10:45 crazycoder hello
10:47 ^7heo moin
10:48 dsdeiz joined #git
10:48 dsdeiz joined #git
10:49 shams joined #git
10:50 Guest11520 left #git
10:50 afuentes joined #git
10:51 xocket joined #git
10:52 didibwoy joined #git
10:56 crazycoder guys we do PR on dev branch....then when we need to do a release we tag it and then should we create a PR to merge the DEV into MASTER or a simple git merge ?
10:56 didibwoy joined #git
10:56 crazycoder ^7heo, ^^
10:58 ^7heo the concept of PR is more about github than git
10:58 ^7heo personally, without github, here's how I would do the thing:
10:58 ^7heo 1. master contains the latest stable
10:58 venmx joined #git
10:58 ^7heo 2. dev contains the dev branch
10:59 ^7heo 3. you might have feature branches on top of dev at any time
10:59 ^7heo 4. when you get a dev you wanna relase, tag it with the release
10:59 thebope_ joined #git
10:59 ^7heo 5. then checkout master, git rebase $tag
10:59 ^7heo and push
11:00 ^7heo (you can git merge also, with fast forward it does not matter)
11:00 ngrin_ joined #git
11:00 zulutango joined #git
11:00 dario_ joined #git
11:01 crazycoder i am reading
11:01 benedikt joined #git
11:01 julius_ joined #git
11:02 benedikt what is the hash at the end of the output from `git describe --tags`? It doesn't match any commit id
11:02 crazycoder ^7heo, ok it makes sense but i have a doubt...
11:02 InfoTest joined #git
11:02 crazycoder we need to generate a changelogs
11:02 crazycoder so we would like to show all the merge requests we did between specific releases
11:02 crazycoder at the moment we do git log --pretty=format:"%h; %s" v2.0.1..v2.0.2
11:03 obiwahn how would i best open a file and it's counterpart from some other revision at the same time in vim?
11:03 ^7heo crazycoder: would work yes.
11:03 crazycoder and if we add --merges we can see the merges instead of just commits
11:03 obiwahn the other rev can be read only obviously
11:03 crazycoder ^7heo, ok but if i do a rebase how can see it on logs ?
11:04 crazycoder does it will be a "merge" ?
11:04 ^7heo crazycoder: if you fast-forward, by default, you won't have a merge commit, afaik
11:04 crazycoder --merges
11:04 crazycoder i do
11:04 crazycoder git merge --no-ff dev (when i am on master)
11:04 atomi joined #git
11:04 ^7heo that's weird, I mean, I have a rebase based workflow, because I work alone
11:04 ^7heo but I would expect merge not to create merge commits if there's no merge.
11:04 ^7heo aaah
11:05 Rodya_ joined #git
11:05 ^7heo it's not *by default* with --noo-ff
11:05 ^7heo --no-ff even
11:05 ^7heo it's more like "please explicitely create a merge commit"
11:05 Curly060 joined #git
11:05 crazycoder yes
11:05 ^7heo so yeah.
11:05 crazycoder it create a merge commit
11:05 crazycoder with all
11:05 obiwahn And what is your work-flow if you have to backport changes from a range of commits (that contain not only the changes you want) to a released branch?
11:06 crazycoder so the thing is create a merge commit with a specific comment
11:06 crazycoder and then to create the changelogs i need --merges option
11:06 crazycoder in this way i will only see the merge
11:06 crazycoder right ?
11:07 xocket joined #git
11:09 crazycoder ^7heo, but wait
11:09 crazycoder we are two devs
11:09 crazycoder if one dev push to dev specific things
11:09 crazycoder and i do the same
11:09 crazycoder the problem is when i do git merge --no-ff dev (into master)
11:09 crazycoder becuase the dev has my commits and his commits
11:09 crazycoder so when i have to put a comment on git merge ..
11:10 crazycoder what comment should i add? becasuse there are many things
11:11 ^7heo wait wat?
11:11 crazycoder i mean... when i do git merge --no-ff dev into master i should add comments for many commits
11:11 pbrewczynski joined #git
11:12 crazycoder what about each dev add his merges and then i do the last on master?
11:15 crazycoder ^7heo, devs are doing their merge into their dev branches...then they push their dev on the "official" dev....then there will be the last merge of dev into master
11:15 mischat joined #git
11:15 crazycoder right ^7heo ^?
11:16 josuebrunel joined #git
11:17 ^7heo crazycoder: well I dunno about your workflow but it sounds alright
11:18 djb-irc joined #git
11:18 josuebrunel left #git
11:20 texinwien joined #git
11:22 Curly060 left #git
11:24 InfoTest joined #git
11:24 moei joined #git
11:27 a_thakur_ joined #git
11:28 xocket joined #git
11:30 leeN joined #git
11:31 kuldeep joined #git
11:32 freimatz joined #git
11:38 rbern joined #git
11:39 JeroenT joined #git
11:44 qt-x joined #git
11:51 Glooomy joined #git
11:51 jbitdrop joined #git
11:53 cdown joined #git
11:56 somebody joined #git
11:56 crazycoder ^7heo, still here?
11:56 jordila joined #git
11:56 crazycoder why did you say
11:57 crazycoder git checkout master
11:57 crazycoder git rebase $tag
11:57 crazycoder git push
11:57 crazycoder why rebase?
11:57 * jordila has just learned to 'git checkout' not only branches but commits and files too...
11:58 InfoTest joined #git
11:58 cqi joined #git
12:02 davimore joined #git
12:05 Rodya_ joined #git
12:06 paladinn joined #git
12:10 re1 joined #git
12:11 anbz joined #git
12:13 LukasT joined #git
12:14 goTAN joined #git
12:19 Sergiu joined #git
12:21 cbreak jordila: most git commands have two, three or more uses :)
12:22 _ikke_ I see it as modifiers of a common behaviour
12:22 * jordila loves free (as in freedom) software in general and git in particular
12:23 _ikke_ If each if these modified behaviours was a single command, the amount of commands would explode, and we would run out of descriptive names to call them
12:23 _ikke_ Not that I say the current situation is ideal
12:24 jaguarmagenta joined #git
12:25 ^7heo crazycoder: rebase is my workflow, so I used rebase
12:26 ^7heo crazycoder: by force of habit
12:26 crazycoder ok but can i not only do
12:26 crazycoder git checkout $tag
12:26 crazycoder and then
12:26 crazycoder push
12:26 crazycoder ?
12:26 ^7heo well, that depends.
12:26 ^7heo if your tag is ff-able on master, you can.
12:26 alhariel joined #git
12:26 TheWild joined #git
12:27 ^7heo if not... it's gonna be rejected
12:28 crazycoder what do you mean with "ff-able on master" ?
12:28 a3Dman joined #git
12:30 adymitruk joined #git
12:30 ^7heo crazycoder: that the path master..tag is fast-forwardable.
12:30 thebope_ joined #git
12:30 crazycoder ok
12:30 skunkz joined #git
12:32 xall joined #git
12:33 crazycoder ^7heo, however i create the tag on dev
12:33 crazycoder i think it is wrong
12:34 rominronin joined #git
12:35 ^7heo a tag is a tag
12:35 ^7heo it doesn't belong to a branch
12:35 ugjka joined #git
12:36 crazycoder ^7heo, so maybe i can move the master doing git reset --HARD $tag what do you think?
12:36 ^7heo to move the master?
12:37 crazycoder yes
12:37 ^7heo why not cd ..; rm -rf $repo?
12:37 ^7heo and then recreate it? :P
12:37 tobiasvl what are you guys talking about
12:37 crazycoder :D
12:37 crazycoder http://stackoverflow.com/questions/6872223/h​ow-do-i-revert-master-branch-to-a-tag-in-git
12:37 ^7heo tobiasvl: he wants to fast-forward his master to a tag
12:37 ^7heo crazycoder: not revert, fast-forward
12:37 ^7heo that's pretty different.
12:38 crazycoder ^7heo, ok i will do git merge --no-ff dev -m "Release v2.0.10"
12:38 crazycoder :)
12:38 ^7heo there's quite a lot wrong about this
12:39 ^7heo but I don't have to time to explain
12:42 LucaTM joined #git
12:42 crazycoder ^7heo, why? http://nvie.com/posts/a-successful-git-br​anching-model/#finishing-a-release-branch
12:42 zeroed joined #git
12:42 crazycoder it is explained here...
12:42 crazycoder i also need a release branch
12:42 crazycoder and then merge this release branch on master with
12:42 crazycoder -no-ff
12:43 acetakwas joined #git
12:43 ^7heo crazycoder: that is an utterly overengineered workflow.
12:43 ^7heo crazycoder: that will always work, the same way as bruteforce will always work.
12:44 byte512 joined #git
12:44 crazycoder ok
12:45 venmx joined #git
12:45 Glooomy joined #git
12:46 maxb Wow, this conversation has been going on for a while :-/  I can't tell whether it is nearing a conclusion or is just generating more questions
12:47 selckin many of those everything looks like a nail blog posts
12:47 grayjoc joined #git
12:51 Junior joined #git
12:52 Cavallari joined #git
12:52 g_zen joined #git
12:53 a3Dman joined #git
12:56 didibwoy joined #git
12:56 thebope_ joined #git
12:58 zeroed joined #git
12:58 zeroed joined #git
12:58 rominronin joined #git
12:59 finalbeta joined #git
12:59 zeroed joined #git
13:02 sgtbigman joined #git
13:03 kazantip joined #git
13:06 Rodya_ joined #git
13:06 danslo joined #git
13:08 a_thakur joined #git
13:08 BadSnappa joined #git
13:13 dario__ joined #git
13:15 mischat joined #git
13:21 rominronin joined #git
13:21 rominronin joined #git
13:23 Flowinho joined #git
13:27 sdothum joined #git
13:28 mindriot101 joined #git
13:30 tvw joined #git
13:30 mischat joined #git
13:30 ngperfex joined #git
13:31 mischat_ joined #git
13:31 Macaveli joined #git
13:33 re1 joined #git
13:34 raijin joined #git
13:35 rominronin joined #git
13:38 Cavallari1 joined #git
13:41 LukasT joined #git
13:45 mindriot101 joined #git
13:47 somebody_ joined #git
13:49 jaguarmagenta joined #git
13:49 JeroenT joined #git
13:49 dsop joined #git
13:49 xocket joined #git
13:49 InfoTest joined #git
13:51 dan2k3k4 joined #git
13:52 Dragonbird joined #git
13:52 xock_stream joined #git
13:55 InfoTest joined #git
13:57 Learath2 joined #git
13:58 rnsanchez_ joined #git
13:59 g_zen hey guys  i  have a project with uses a few git repositories
13:59 g_zen when i push my project
13:59 g_zen and then pull it from another place
13:59 g_zen those git repositories are empty folders
14:00 g_zen what can i do about that?
14:00 tobiasvl you mean when you clone them?
14:00 selckin you go to those reposiitories and do the same
14:00 tobiasvl g_zen: git pull --recurse-submodules
14:00 tobiasvl oh, no
14:00 tobiasvl lol that doesn't do what you want
14:01 tobiasvl git submodule update --init --recursive
14:01 tobiasvl submodule commands are a bit confusing
14:01 tobiasvl g_zen: do you want to automate it? is this a deployment thing?
14:01 g_zen i get this: No submodule mapping found in .gitmodules for path '...'
14:01 g_zen well
14:01 g_zen preferrably automate it
14:01 phanimahesh joined #git
14:02 g_zen not deployment though
14:02 g_zen but i have faced this issue quite a few times
14:02 g_zen when i accidentaly find out that some of my files are just missing
14:03 re1 joined #git
14:03 goTAN joined #git
14:04 phanimahesh joined #git
14:06 g_zen even the folders which contained a git repository dont have the .git folder
14:06 g_zen so supposedly i'd have to clone them again
14:06 g_zen and what if i had changed something?
14:07 selckin getting closer and closer to sounding like the ramblings of a mad man
14:07 Rodya_ joined #git
14:07 tobiasvl g_zen: it sounds like you're using !submodules but don't know how submodules work
14:07 gitinfo g_zen: [!submodule] git-submodule is ideal to add subsidiary git repositories to a git superproject when you do not control the subprojects or more specifically wish to fix the subproject at a specific revision even as the subproject changes upstream. See http://www.git-scm.com/boo​k/en/Git-Tools-Submodules
14:08 g_zen so instead of doing git clone ...... inside my project  i should say git submodule add ......
14:08 g_zen ?
14:09 moritz instead of trying random stuff, you should read some tutorials and documentation
14:09 zeroed joined #git
14:10 kedare joined #git
14:10 g_zen i'll take that as a  yes :D
14:11 moritz that's not what I said :-)
14:12 g_zen yeah well it also didnt work .. :)
14:13 nettoweb joined #git
14:15 maroloccio joined #git
14:15 maroloccio how can i get the commit id of the "last" commit in this linear history?
14:15 tobiasvl maroloccio: git rev-parse HEAD
14:16 mindriot101 joined #git
14:16 tobiasvl or just look at git log
14:16 dreaddymck joined #git
14:16 maroloccio tobiasvl: thanks. sorry, i explained myself wrongly.
14:16 maroloccio tobiasvl: the "other end", not the "tip", or "HEAD"
14:16 tobiasvl so the FIRST commit?
14:16 maroloccio yeah
14:16 finalbeta joined #git
14:16 JeroenT joined #git
14:17 Dragonbird you can say `git rev-list --max-parents=0 HEAD`
14:19 Qria joined #git
14:19 sgtbigman Neat, thanks Dragonbird
14:21 LukasT joined #git
14:21 maroloccio wow, that's so cool, Dragonbird
14:21 wizonesolutions joined #git
14:21 maroloccio but it will actually return other commits if they were git branched using --orphan, right?
14:21 maroloccio no, it won't because it starts from HEAD, right?
14:21 maroloccio just making sure
14:22 kyan joined #git
14:22 maroloccio it walks back from HEAD and output all the "start of history" commits.. so.. let me think.. it would actually find two is there was a merge of two so-called _unrelated histories_, right?
14:22 Dragonbird I believe it will return commits from orphan branches as well
14:22 ghostlight joined #git
14:23 Dragonbird technically it should
14:23 Dragonbird I'll have to read the man page to make sure
14:23 maroloccio --allow-unrelated-histories
14:23 maroloccio i am talking about C (HEAD) being a commit with 2 parents: A and B, both "start of histories"
14:24 maroloccio but i said "linear history", so that does not apply in that particular case
14:25 xall joined #git
14:26 Dragonbird maroloccio: Yep, it will give you orphan commits as well
14:26 Dragonbird Just tried
14:27 jaguarmagenta joined #git
14:29 lektrik joined #git
14:29 shinnya joined #git
14:30 mizu_no_oto joined #git
14:30 jaguarmagenta joined #git
14:31 BackEndCoder joined #git
14:32 todd_dsm joined #git
14:33 chardan joined #git
14:33 JamesBaxter joined #git
14:33 maroloccio Dragonbird: how did you try? i seem not to get the orphans? (in a _linear_ history, no merging)
14:33 maroloccio e.g.: cd "$(mktemp -d)"; git init; touch a; git checkout -b a; git add a; git commit -m.; git checkout --orphan b; touch b; git add b; git commit -m.; git rev-list --max-parents=0 HEAD; git log --graph --all --oneline --decorate=full
14:33 jbitdrop joined #git
14:34 maroloccio i don't get the a branch commit from the HEAD of the b branch (i am not getting "all starts of histories", just the start of that b history)
14:34 maroloccio Dragonbird: confirm?
14:35 Dragonbird I did a quick orphan checkout on existing project of mine. let me see
14:35 Dragonbird hang on
14:35 davix[matrix] joined #git
14:36 ShekharReddy joined #git
14:37 adac I checked out a branch (A) from another branch (B) and did some changes in Branch A. Now I wanted to again merge changes from Branch B into Branch A. Is that possible, and if yes, how?
14:37 ToxicFrog git checkout A; git merge B?
14:39 Kestrel-029 joined #git
14:40 adac ToxicFrog, git merge module/mosquito_traps
14:40 adac merge: module/mosquito_traps - not something we can merge
14:40 wizonesolutions joined #git
14:41 ToxicFrog adac: is that the entire output?
14:41 LukasT joined #git
14:41 ToxicFrog what does "git log module/mosquito_traps" show? "git show module/mosquito_traps"?
14:41 nettoweb joined #git
14:42 adac ToxicFrog, https://gist.github.com/anonymous​/475d1a763e5519ba32dffc5027cbdad0
14:43 adac ToxicFrog, ok I had to add origin: origin/module/mosquito_traps
14:43 adac but why?
14:44 ToxicFrog Presumably because you don't have a local branch named "module/mosquito_traps"
14:44 ToxicFrog try "git branch -a"
14:44 patrickr[m] joined #git
14:44 kineticrootkit[m joined #git
14:44 unclechu joined #git
14:44 M-nated joined #git
14:44 WillianPaiva[m] joined #git
14:44 saintaquinas[m] joined #git
14:44 kikijiki[m] joined #git
14:44 samwise[m] joined #git
14:44 M-meznak joined #git
14:44 M-shine joined #git
14:44 Remramm joined #git
14:44 hallogallo[m] joined #git
14:44 mtso[m] joined #git
14:44 M-mistake joined #git
14:44 monomon[m] joined #git
14:44 srkrishna[m] joined #git
14:44 Dhanya[m] joined #git
14:44 musicmatze[m] joined #git
14:44 aviraldg joined #git
14:44 Ineentho[m] joined #git
14:44 alaeri[m] joined #git
14:44 warrantyvoid joined #git
14:44 LukasT joined #git
14:44 Tiaan[m] joined #git
14:44 timlyo[m] joined #git
14:44 kjsage[m] joined #git
14:44 Carlos[m] joined #git
14:44 M64137[m] joined #git
14:44 eltopo[m] joined #git
14:45 ToxicFrog Some commands (including "checkout") will automatically look for a remote branch if there isn't a local one with the name you specified (and in some cases will automatically create such a local branch, IIRC); merge isn't one of them.
14:45 adac ToxicFrog, yes it is a remote one. But how can I have it also locally? Shouldnt that be done by git pull?
14:45 Miika-- left #git
14:45 adac ToxicFrog, ok that explains it
14:45 adac so only with checkout I will have it locally
14:45 adac I guess
14:48 krillr joined #git
14:48 Karazhan joined #git
14:48 vuoto joined #git
14:49 finalbeta joined #git
14:49 ToxicFrog adac: so, what you have there is a "remote tracking branch". It's stored locally (i.e. accessing it doesn't require a net connection), but is an exact copy of the branch as it existed on the remote last time you fetched.
14:50 adac ToxicFrog, I see
14:50 ToxicFrog Because of that, it's not meant to be committed to directly, although you can merge or cherry-pick *from* it into local branches.
14:50 adac thanks for the explanation ToxicFrog!
14:51 ToxicFrog If you wanted to work "on that branch", you'd create a local copy of it (deliberately or automatically), make your commits on the local copy, and then "git push" to update the remote branch (and your local copy of the remote branch) with your new commits.
14:51 BackEndCoder joined #git
14:51 ToxicFrog You're welcome :)
14:51 seishun joined #git
14:52 vuoto joined #git
14:52 steelsw0rd joined #git
14:53 xocket joined #git
14:54 isysd joined #git
14:56 Noldorin joined #git
14:56 venmx joined #git
14:56 AaronMT joined #git
14:57 Dragonbird maroloccio: sorry. hastily I made a mistake. You are correct it won't show orphan commits
14:57 garo joined #git
14:58 julius_ joined #git
14:59 Starky joined #git
15:00 Anticom joined #git
15:00 vuoto joined #git
15:00 Anticom Hi all. How would i supply multipe pathspecs to 'git submodule init' ?
15:01 Anticom is it a semicolon-seperated list?
15:02 gargola joined #git
15:02 eivarv joined #git
15:02 JamesBaxter joined #git
15:06 Darren_ joined #git
15:06 venmx joined #git
15:07 Gsham joined #git
15:08 Rodya_ joined #git
15:08 al-damiri joined #git
15:09 vuoto joined #git
15:11 jaguarmagenta joined #git
15:11 garo Situation: I have computers A and B that both have a dir /home/garo/scripts that I want to track in git. I have 5 scripts that i use on both A and B but both systems also have a couple of unique scripts.
15:11 garo Is it possible to track both dirs in a way that I can push changes to the shared scripts from system A to system B without pushing the unique (but also tracked by git) scripts ?
15:12 Noldorin joined #git
15:13 ToBeCloud joined #git
15:13 Anticom garo: why not move the unique scripts to a subfolder 'machine-specific' etc. and track those wich unique names aswell. Then have a symlink in /home/garo/scripts which points to the desired unique script and let git ignore the symlink
15:13 mischat joined #git
15:14 Anticom that way you'd have to create the symlink after checking out the repo but you get what you where asking for :)
15:17 garo seems like a ugly hack...
15:17 davimore joined #git
15:18 vuoto joined #git
15:18 BackEndCoder joined #git
15:18 a_thakur joined #git
15:19 vuoto joined #git
15:23 venmx joined #git
15:23 gamester joined #git
15:23 jimi_ joined #git
15:24 nettoweb joined #git
15:24 seishun joined #git
15:25 subhojit777 joined #git
15:29 Anticom garo: well it's the first thing that came to my mind. You could of course also manage your unique scripts in different branches of the same repo.
15:29 rominronin joined #git
15:29 mischat joined #git
15:33 dendazen joined #git
15:35 LukasT joined #git
15:37 LukasT joined #git
15:40 lohfu joined #git
15:42 ayogi joined #git
15:42 dario_ joined #git
15:43 ochorocho__ joined #git
15:44 BackEndCoder joined #git
15:44 dreiss joined #git
15:45 mischat__ joined #git
15:47 dario__ joined #git
15:47 Cavallari joined #git
15:48 ayjay_t joined #git
15:52 xock_stream joined #git
15:52 goTAN joined #git
15:52 lektrik joined #git
15:58 LukasT joined #git
16:00 UniFreak joined #git
16:00 JamesBaxter joined #git
16:01 a3Dman joined #git
16:01 UniFreak joined #git
16:03 UniFreak joined #git
16:04 Rickmasta joined #git
16:04 ochorocho__1 joined #git
16:04 UniFreak joined #git
16:06 mischat joined #git
16:06 oskarkv joined #git
16:07 mindriot101 joined #git
16:07 NeverDie joined #git
16:07 King_Hual joined #git
16:07 King_Hual joined #git
16:08 LukasT joined #git
16:08 Rodya_ joined #git
16:08 m___ joined #git
16:11 ojdo joined #git
16:12 Vinnie_win_x joined #git
16:12 EricPoe joined #git
16:13 danslo_ joined #git
16:14 stomplee joined #git
16:15 julius_ joined #git
16:15 zacts joined #git
16:18 Vinnie_win joined #git
16:20 dsdeiz joined #git
16:20 dsdeiz joined #git
16:22 lindenle joined #git
16:23 irco joined #git
16:25 blackwind_123 joined #git
16:27 rominronin joined #git
16:31 rominronin joined #git
16:31 xocket joined #git
16:32 Rickmasta joined #git
16:33 xock_stream joined #git
16:33 flaguy48 joined #git
16:36 navidr joined #git
16:39 einnjo joined #git
16:41 GavinMagnus joined #git
16:42 eivarv joined #git
16:42 NeXTSUN joined #git
16:43 Dragonbird joined #git
16:44 jimi_ joined #git
16:46 finalbeta joined #git
16:48 Phylock joined #git
16:50 courrier joined #git
16:53 raspado joined #git
16:54 JeroenT_ joined #git
16:57 julius_ joined #git
16:57 xall joined #git
17:00 GavinMagnus left #git
17:00 anuxivm joined #git
17:02 kbs joined #git
17:02 A5101 joined #git
17:07 bket joined #git
17:08 Domhack joined #git
17:09 Rodya_ joined #git
17:12 JeroenT joined #git
17:12 acetakwas joined #git
17:12 jaguarmagenta joined #git
17:16 seishun joined #git
17:16 dreiss joined #git
17:18 Learath2 joined #git
17:22 Charliechin joined #git
17:23 mischat joined #git
17:24 NeverDie joined #git
17:25 crazycoder joined #git
17:31 Flowinho joined #git
17:33 sea-gull_ joined #git
17:33 raijin joined #git
17:33 rominronin joined #git
17:35 Siegfried joined #git
17:36 jccn joined #git
17:37 meauses joined #git
17:40 acetakwas joined #git
17:42 goTAN joined #git
17:42 Domhack joined #git
17:42 refried_ joined #git
17:42 grayjoc joined #git
17:45 cdg joined #git
17:46 rath joined #git
17:47 mischat_ joined #git
17:48 dikiaap joined #git
17:51 lordjancso joined #git
17:54 nidr0x joined #git
17:55 cagedwisdom joined #git
17:56 kab joined #git
17:57 manuel joined #git
17:57 breezy_ joined #git
17:59 acetakwas joined #git
18:00 manuel joined #git
18:02 tang^ joined #git
18:03 Goplat joined #git
18:06 qws-user-1228 joined #git
18:08 julius_ joined #git
18:10 NeverDie joined #git
18:10 Rodya_ joined #git
18:14 anonymuse joined #git
18:18 inflames joined #git
18:19 JeroenT joined #git
18:19 jimi_ joined #git
18:23 dsdeiz joined #git
18:23 dsdeiz joined #git
18:23 JyZyXEL left #git
18:24 anonymuse joined #git
18:27 fragMental joined #git
18:32 Mobutils joined #git
18:34 manuel joined #git
18:36 overlord_tm joined #git
18:39 anonymuse joined #git
18:45 venmx joined #git
18:49 manuel joined #git
18:49 Akimb joined #git
18:51 seishun joined #git
18:51 mindriot101 joined #git
18:52 as_ joined #git
18:53 as_ left #git
18:54 atomi joined #git
18:55 nettoweb joined #git
18:56 d0nn1e joined #git
19:00 therealGent joined #git
19:02 satifant joined #git
19:02 agileadam joined #git
19:06 anuxivm joined #git
19:11 Rodya_ joined #git
19:12 Learath2 joined #git
19:12 xall joined #git
19:13 jaguarmagenta joined #git
19:13 goTAN joined #git
19:14 brent__ joined #git
19:15 kulelu88 joined #git
19:17 A5101 joined #git
19:22 SwiftMatt joined #git
19:23 realz joined #git
19:28 timwis Morning folks, does anyone know what ':' means in a git diff? (as opposed to ++, --, -->) (as seen on http://sweltering-neck.surge.sh/ about 25% the way down, the yellow rows)
19:29 einnjo joined #git
19:29 dario__ joined #git
19:29 BlaXpirit timwis, is this a merge?
19:30 BlaXpirit git diff doesn't produce ':'
19:30 seishun joined #git
19:30 moritz this does not look like a git diff
19:30 moritz git diff also doesn't produce →
19:30 NeverDie joined #git
19:30 irco joined #git
19:32 nwkj86 joined #git
19:33 timwis Ah, it's actually using daff (https://github.com/paulfitz/daff), but I thought daff used a standard diff format that git also used, so I thought it safe to ask here :-/
19:33 timwis I'm trying to compare two CSV files
19:34 BlaXpirit so  not a merge?  then this makes no sense
19:34 Macaveli joined #git
19:34 timwis well, the purple cells indicate changes a la git diff --word-diff
19:35 rominronin joined #git
19:35 Masber joined #git
19:35 paladinn joined #git
19:35 paladinn joined #git
19:37 mindriot101 joined #git
19:38 spacemud joined #git
19:39 meauses joined #git
19:41 adac joined #git
19:41 aard_ joined #git
19:41 xall joined #git
19:42 mindriot101 joined #git
19:42 mischat joined #git
19:45 multi_io_ joined #git
19:47 govg joined #git
19:48 Flowinho joined #git
19:49 manuel joined #git
19:49 barduck joined #git
19:49 manuel joined #git
19:50 timwis Is there a way to use git diff --word-diff to use commas as word separators instead of spaces?
19:50 timwis Not having much luck with --word-diff-regex..
19:50 Gsham joined #git
19:52 terens joined #git
19:52 terens hello
19:52 terens why they call eval(ssh-agent) ?
19:53 nicksloan joined #git
19:55 rorro joined #git
19:55 WayToDoor joined #git
19:55 dreiss joined #git
19:57 xocket joined #git
19:58 railssmith joined #git
20:00 nurey joined #git
20:01 xocket joined #git
20:02 nurey joined #git
20:02 moritz you actually want eval $(ssh-agent)
20:03 moritz ssh-agent starts a background process, and prints some environment variables
20:03 moritz eval'ing that ouptut is the way to connect your current shell with that background process
20:04 LucaTM joined #git
20:05 eblip joined #git
20:06 Doginal joined #git
20:07 manuel joined #git
20:09 Siegfried joined #git
20:10 seishun joined #git
20:11 elastix joined #git
20:11 Rodya_ joined #git
20:14 aarobc joined #git
20:14 xock_stream joined #git
20:15 Peng_ joined #git
20:15 Macaveli joined #git
20:17 Seveas these days however, sensible desktops set this up for you and you don't need to run your agent manually
20:18 dminuoso left #git
20:19 dzho it's almost to the point you don't have to think about your crypto at all!
20:19 dzho what a world, what a world.
20:23 dsdeiz joined #git
20:23 dsdeiz joined #git
20:24 elastix joined #git
20:24 Rodya_ joined #git
20:30 kulelu88 joined #git
20:31 axisys joined #git
20:33 lindenle joined #git
20:34 dikiaap joined #git
20:37 stux|RC-only joined #git
20:38 Flowinho joined #git
20:39 jstein_ joined #git
20:39 vdv joined #git
20:40 stux|RC-only joined #git
20:43 raspado joined #git
20:44 skunkz joined #git
20:48 CrashOverride joined #git
20:58 sq-one joined #git
20:59 Zumochi left #git
21:01 dario__ joined #git
21:01 boombatower_ joined #git
21:02 Anja joined #git
21:02 nettoweb joined #git
21:02 Merak joined #git
21:03 anuxivm joined #git
21:04 dodobrain joined #git
21:04 nivram913 joined #git
21:05 Merak I have git version 2.10.2.   git checkout --recursive  gives me submodules which only have a .git directory and no working tree.  I can see a history in them with 'git log'.
21:05 nivram913 left #git
21:06 Merak I guess that means I get bare repositories. From what I read, this is not what I expect.  Is something wrong?
21:06 _ikke_ Hmmm, I never heard of bare submodules
21:07 _ikke_ If there is a .git there, it doesn't look like a bare repository though
21:07 _ikke_ a .git directory
21:08 sea-gull_ joined #git
21:13 afuentes joined #git
21:14 jaguarmagenta joined #git
21:14 aspiers joined #git
21:14 Masber joined #git
21:14 xocket joined #git
21:16 dsdeiz joined #git
21:17 crazycoder joined #git
21:19 nettoweb joined #git
21:23 kazantip joined #git
21:25 goTAN joined #git
21:26 NeverDie joined #git
21:29 vuoto joined #git
21:29 mizu_no_oto joined #git
21:32 WaReZ joined #git
21:37 endiruna joined #git
21:37 re1_ joined #git
21:40 Doginal_ joined #git
21:42 CrashOverride joined #git
21:42 Goplat joined #git
21:42 mizu_no_oto joined #git
21:43 thebope_ joined #git
21:43 Atm0spher1c joined #git
21:50 venmx joined #git
21:50 PCatinean joined #git
21:53 mindriot101 joined #git
21:54 Seveas never heard of bare submodules. What does git ls-tree HEAD say inside one of them?
21:56 jordila joined #git
21:57 cdown joined #git
21:58 NeXTSUN joined #git
22:01 mindriot101 joined #git
22:02 Vampire0_ joined #git
22:03 snthsnth joined #git
22:04 WaReZ joined #git
22:06 tvw joined #git
22:07 re1 joined #git
22:08 therealGent joined #git
22:10 Merak Seveas: git ls-tree .  : "fatal: Not a valid object name ."
22:10 Merak For any who wish to see it themselves.   git clone --retursive https://gitgud.ioKiri/ModPackNoDelayHDX.git
22:11 Merak oops.  'recursive' , not 'retursive'
22:11 Merak (that was not the error i wrote about)
22:12 Seveas Merak: that is not the command I asked you to run...
22:12 Merak Seveas: I ran what you asked also.
22:13 Seveas and what did it say?
22:13 _ikke_ could not resolve..
22:13 _ikke_ missing /
22:14 _ikke_ fatal: unable to connect to gitgud.io:
22:14 _ikke_ gitgud.io[0: 104.238.217.12]: errno=Connection refused
22:14 Merak Seveas: i did make an error running the command you asked.  git ls-tree HEAD   shows a listing of what would be in the working directory.
22:15 xock_stream joined #git
22:15 _ikke_ Merak: Something goes wrong when cloning
22:15 Merak _ikke_: without typos, here is the command: git clone --recursive https://gitgud.io/Kiri/ModPackNoDelayHDX.git
22:16 _ikke_ Merak: Yeah, I manualle fixed the url, but it looks like the .gitmodules contains git://gitgut.io, which is refused
22:16 Merak There is 1 submodule which has a protocol specified which is unreachable.  Yes. that one.  The rest appear to clone fine.
22:17 Merak (from the console output, at least)
22:17 _ikke_ correct\
22:17 _ikke_ So I think that's why it failed to checkout the rest
22:17 Seveas yeah, I agree with that guess
22:17 _ikke_ git submodule foreach 'git checkout -f'
22:18 Seveas Merak: could you try an older git? The submodule code got a rewrite recently, this may be a sideeffect
22:18 Seveas heh, "recently". it started in the 2.7 series...
22:20 Merak Path of least resistance to me seems like fixing the 1 submodule which has the improper protocol.
22:20 Merak and testing again
22:20 Seveas true, but I consider this a bug to be fixed. Would be nice to know if it's a regression
22:20 moei joined #git
22:20 Seveas so please don't fix the repo just yet until I have a chance to test :)
22:20 Merak alright, i will get an older version.
22:22 Seveas building git 2.6 now to test
22:24 Seveas hmm
22:24 Seveas $ ./git clone --recursive https://gitgud.io/Kiri/ModPackNoDelayHDX.git
22:24 Seveas Cloning into 'ModPackNoDelayHDX'...
22:24 Seveas warning: templates not found /home/dennis/share/git-core/templates
22:24 Seveas fatal: Unable to find remote helper for 'https'
22:24 Seveas _ikke_: ever seen that in a freshly built git?
22:24 _ikke_ Seveas: nope
22:24 eivarv joined #git
22:24 _ikke_ I'm using my own built git versions
22:25 _ikke_ But even when I use make && make install I haven't encountered it
22:25 raspado joined #git
22:25 Seveas didn't do make install, trying to run from the build dir
22:25 Seveas PATH=.:$PATH works around it, but that's not supposed to be needed...
22:26 _ikke_ Seveas: That's why I guess
22:26 Seveas not a regression, 2.6.2 also doesn't check out the repos
22:26 _ikke_ ok
22:27 Merak I am building 2.6.6 -- but I guess there is no need for me to try that now.
22:27 Seveas Merak: correct.
22:27 acetakwas joined #git
22:29 NeverDie joined #git
22:32 adac joined #git
22:34 tang^ joined #git
22:36 cyberz hello guys
22:38 Seveas _ikke_: Merak: I just read through the code and our suspicion is correct
22:38 Seveas if *any* submodule clone fails, none get checked out
22:39 cyberz I use quite often the git commit --fixup + git rebase --autosquash trickery and I'm quite happy with it. The only boring thing is that sometimes, If I fixup various commits, I really don't know by heart which one is the oldest, and then I have to grep and find it out manually. Is there anything that could automate that? Ideally I would like to git rebase -i --autosquash --find-the-first-fixup-commit-on-your-own
22:39 Seveas the checking things out part is still in the shell version of the code, so it's a bit tricky to fix. aybe better to wait until that gets ported over.
22:41 mischat joined #git
22:41 isysd joined #git
22:43 CrashOverride joined #git
22:43 astrofog joined #git
22:44 ahmed_elgabri joined #git
22:49 anuxivm joined #git
22:49 Seveas cyberz: 'the oldest' is generally an inefficient question, as git may need to go all the way to the beginning of time
22:52 Seveas this gives you the oldest commit that does not appear in your upstream branch (the one you push to/fetch from) that matches 'fixup!': git log --reverse @{u}.. --grep '^fixup!' --format=%h | head -n1
22:52 Seveas note that you can't use git log -n1 here as the limiting is done before --reverse takes effect
22:52 m0viefreak joined #git
22:53 cyberz Seveas nice thanks!
22:56 Alexandreffaria joined #git
22:57 AfroBomaye[m] joined #git
22:57 GavinMagnus joined #git
22:57 AfroBomaye[m] left #git
23:01 cagedwisdom joined #git
23:02 tvw joined #git
23:03 raijin joined #git
23:06 chid_ joined #git
23:12 fiete joined #git
23:14 goTAN joined #git
23:14 jaguarmagenta joined #git
23:15 xocket joined #git
23:20 finalbeta joined #git
23:22 alex1a joined #git
23:23 madewokherd joined #git
23:26 dreaddymck joined #git
23:29 mellernoia joined #git
23:31 Sasazuka joined #git
23:34 GavinMagnus left #git
23:34 multi_io_ left #git
23:34 multi_io_ joined #git
23:35 multi_io_ left #git
23:35 multi_io joined #git
23:35 multi_io looks like git log --follow never follows files in a git subtree directory into the past beyond the commit where the directory was added?
23:36 Seveas multi_io: was it added with --squash?
23:37 multi_io Seveas: no
23:38 multi_io the whole history of both repos (base and subtree) is there
23:39 Seveas how about git log --full-history --follow?
23:43 multi_io Seveas: nope, doesn't work either
23:45 Seveas rats. Is the repo public?
23:45 multi_io nah. only the subtree is.
23:46 multi_io https://github.com/multi-io/mydocs-pub
23:47 mischat joined #git
23:48 mischat joined #git
23:48 johnville joined #git
23:49 mischat joined #git
23:49 multi_io I have a mydocs-priv with its own history, added the above to it into a subdir "pub" via git subtree add -P pub https://github.com/multi-io/mydocs-pub.git master
23:49 mischat joined #git
23:50 pdgwien joined #git
23:50 multi_io then e.g. "git log --follow -- pub/emacs/elisp/debugging.txt" outputs nothing(??)
23:50 mischat joined #git
23:51 Seveas that may be normal due to history simplification, but --full-history (before the --, maybe before the --follow) should prevent that
23:51 multi_io "git log --full-history -- pub/emacs/elisp/debugging.txt" outputs the history back to the merge, not before that
23:51 mischat joined #git
23:51 Seveas you need to combine the two
23:52 multi_io "--follow --full-history"  -- again no output
23:52 mischat joined #git
23:52 Seveas ok, that's confusing to say the least
23:53 Seveas would need to reproduce it to say something sensible, but that's not going to happen right now.
23:53 multi_io I guess you should be able to reproduce with an empty / test repository that you subtree add the above one to
23:53 mischat joined #git
23:54 mischat joined #git
23:55 Seveas hmm
23:55 Seveas after adding that to a random non-empty repo, all git-log invocations I'm trying end up empty
23:56 Seveas ah
23:57 Seveas git log -- emacs/elisp/debugging.txt
23:57 Seveas the rename is done in the dirty merge done by the subtree code
23:58 multi_io yeah, that renames it from emacs/elisp/debugging.txt to pub/emacs/elisp/debugging.txt
23:58 multi_io necessarily
23:58 multi_io hence I thought --follow should follow the rename :)
23:59 multi_io you called the "git log -- emacs/elisp/debugging.txt" from the pub dir (i.e. the subdir), right?

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