Perl 6 - the future is here, just unevenly distributed

IRC log for #git, 2016-07-15

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

All times shown according to UTC.

Time Nick Message
00:00 nettoweb joined #git
00:03 f3r70rr35f joined #git
00:03 devbug joined #git
00:03 Peuc joined #git
00:04 Darren_ joined #git
00:04 xall_ joined #git
00:09 inflames joined #git
00:10 piling_ joined #git
00:10 dilijev joined #git
00:10 knittl joined #git
00:10 mika joined #git
00:10 fstd_ joined #git
00:11 dilijev left #git
00:12 zacts joined #git
00:13 dilijev_ joined #git
00:20 Orion3k Hi, anyone know how to get a post-receive hook to work in a bare directory when pushing to a remote with a branch and not master?  I am getting this error - remote: fatal: You are on a branch yet to be born
00:20 romerocesar joined #git
00:22 tristanp joined #git
00:23 mwhooker joined #git
00:32 Literphor joined #git
00:40 Orion3k nevermind, i forgot the branch name on the checkout command in the post-receive hook
00:40 romerocesar joined #git
00:41 codebam so I have a 3.4.10 kernel, how can I merge 3.4.12?
00:42 Chiko joined #git
00:42 codebam they are tags
00:43 osse codebam: git merge <thing>
00:43 codebam but its a tag, how do I merge tags?
00:44 osse codebam: git merge <tag>
00:44 f3r70rr35f joined #git
00:45 codebam ohh okay
00:45 codebam thanks lol
00:45 codebam so I would add it as a remote right, like git remote add upstream <url>
00:45 codebam and then git merge upstream/tag?
00:46 rominronin joined #git
00:46 osse fetch first
00:46 osse and just git merge tag
00:46 osse because tags are global.
00:46 osse there is no tag versus upstream/tag like with branches
00:47 cqi joined #git
00:52 dx joined #git
00:52 xlegoman joined #git
00:55 codebam oh okay
00:57 demimeta joined #git
00:59 tristanp joined #git
01:00 romerocesar joined #git
01:03 dilijev_ hello!
01:03 gitinfo dilijev_: hi! I'd like to automatically welcome you to #git, a place full of helpful gits. Got a question? Just ask it — chances are someone will answer fairly soon. The topic has links with more information about git and this channel. NB. it can't hurt to do a backup (type !backup for help) before trying things out, especially if they involve dangerous keywords such as --hard, clean, --force/-f, rm and so on.
01:04 dilijev_ !backup
01:04 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
01:04 dilijev_ Ok looks like I can send messages to this channel now thanks to nickserv
01:05 dilijev_ I have a question. Do git --pretty="" format strings support date formatting? AFAICT they do not, and I've resorted to using backspace characters to chop off the parts I don't care about
01:06 osse dilijev_: there are a couple of different formats you can pick from, but nothing akin to strftime()
01:07 osse dilijev_: never mind, you can do exactly that :P
01:07 xall_ joined #git
01:07 osse dilijev_: --date=format:%c
01:07 dilijev_ oh
01:07 osse dilijev_: then use the --pretty thingies that says "respects --date option"
01:08 dilijev_ ok how can i combine that with [pretty] format in the .gitcongfig?
01:08 dilijev_ so i guess I'll need to update my aliases to take the --date option as well as the date pretty format string?
01:08 osse i suppose so
01:08 dilijev_ is there an eruivalent [date] .gitconfig option so I can refer to those formats by name?
01:09 diprotic joined #git
01:09 osse dilijev_: ehh, nope: git config log.date '%c blabla'
01:09 dilijev_ equivalent*
01:09 dilijev_ oh that's even better
01:09 dilijev_ i think
01:09 netj joined #git
01:09 dilijev_ i don't usually care about more than one date format
01:09 ome joined #git
01:13 dilijev_ Using git version 2.8.2.windows.1
01:13 dilijev_ git log --date=<anything>
01:13 dilijev_ fatal: unknown date format
01:13 dilijev_ any ideas?
01:13 dilijev_ osse: (tagging you)
01:13 dilijev_ Sorry; I'm new to IRC
01:13 osse what did you try
01:14 dilijev_ git log --date=short
01:14 dilijev_ fatal: unknown date format
01:14 dilijev_ http://linux.die.net/man/1/git-log
01:14 osse try rfc
01:15 pks joined #git
01:15 dilijev_ git log --date=rfc
01:15 dilijev_ fatal: unknown date format
01:15 osse well then
01:15 osse i have no clue
01:15 dilijev_ osse: user error
01:16 dilijev_ osse: I had log.date set incorrectly in .gitconfig
01:16 osse heh
01:16 dilijev_ it was using that first and failing
01:16 osse i would think --date given on the CLI would override that, but...
01:17 dilijev_ osse: apparently no?
01:17 osse apparently
01:17 dilijev_ maybe if the value in log.date was correct in the first place?
01:17 dilijev_ I set log.date=rfc and then did git log --date=short and it worked
01:18 dilijev_ i guess it validates the .gitconfig and then defers to command line arg first. that seems half broken
01:18 osse usually CLI overrides the config files
01:18 osse so this is unexpected to me
01:18 dilijev_ right. it did so. but only when the config file was valid
01:19 dilijev_ from log given the above: [ a956dde ]  2016-07-11
01:19 dilijev_ it used short from the CLI and not the rfc from the gitconfig
01:19 dilijev_ short is generally the format I prefer
01:19 dilijev_ but even so
01:20 MattMaker joined #git
01:20 dilijev_ how do I use strftime()-like formats?
01:20 osse like i said
01:20 osse --date=format:...
01:20 dilijev_ looks like it only takes these:{relative,local,default,iso,rfc,short,raw}
01:21 osse maybe it's a new feature
01:21 dilijev_ osse: thanks, it works when I use the right flags
01:21 dilijev_ git log --date=format:"%Y-%m-%d"
01:22 dilijev_ I was expecting to write something like yyyy-MM-dd but this works, too
01:22 osse that doesn't work with strftime either
01:23 PaulCapestany joined #git
01:23 dilijev_ right -- in truth I've never actually used that API and I assumed incorrectly about it's format. But this style of formatting makes more sense from a c API perspective
01:23 dilijev_ osse: Thanks for your help in showing me how to do this; it had been bothering me for months :P
01:24 eizua joined #git
01:24 fahadash joined #git
01:25 Guest39441 joined #git
01:27 dsantiago joined #git
01:27 relgar joined #git
01:29 nettoweb joined #git
01:30 dreiss joined #git
01:31 romerocesar joined #git
01:32 Ph_ joined #git
01:33 cyan__ joined #git
01:38 david007 joined #git
01:41 romerocesar joined #git
01:43 meloc joined #git
01:46 tristanp joined #git
01:46 meloc joined #git
01:48 ilbot3 joined #git
01:48 Topic for #git is now Welcome to #git, the place for git help and surveys | Public logs at http://goo.gl/BuUi5o | Current stable version: 2.9.1 | First visit? Read: http://jk.gs/git | Getting "cannot send to channel"? /msg gitinfo .voice | rerere: the git that keeps on giving
01:49 stantonk joined #git
01:50 unreal joined #git
01:50 NimeshNeema joined #git
01:52 shaggycat joined #git
01:53 thiago joined #git
01:53 sangy joined #git
01:54 xall_ joined #git
01:55 armenzg joined #git
01:57 justanotheruser joined #git
01:57 BombStrike joined #git
01:58 martinium joined #git
01:58 l4v2 joined #git
01:59 tcsc joined #git
01:59 Dougie187 joined #git
01:59 f3r70rr35f joined #git
02:05 drbean joined #git
02:07 p4trix joined #git
02:07 p4trix Is there a channel to talk about bitbucket pipeline?
02:08 harish_ joined #git
02:09 Goplat joined #git
02:12 romerocesar joined #git
02:14 Meow-J joined #git
02:16 MattMaker joined #git
02:16 Guest39441 joined #git
02:17 tristanp joined #git
02:18 danimal2 joined #git
02:19 Noldorin joined #git
02:19 hexagoxel joined #git
02:19 netj joined #git
02:22 Ankhers joined #git
02:23 Masterphi joined #git
02:27 durham_ joined #git
02:28 Guest39441 joined #git
02:31 romerocesar joined #git
02:32 Literphor joined #git
02:37 f3r70rr35f joined #git
02:38 Limes_ joined #git
02:42 Lyqyd joined #git
02:42 lotus joined #git
02:42 lotus Hey, how can I leave a single file unmerged?
02:43 lotus I want to pull in all of the changes on a branch, but one of the plugins is conflicting with my current development.  I'd like to keep my plugin file in-tact but leave the option to resolve the conflicts for another time.
02:45 Guest39441 joined #git
02:48 rominronin joined #git
02:48 thiago lotus: ok, so don't fix the merge
02:48 thiago you won't be able to commit the merge until you have, though
02:49 thiago just disable the plugin in your tree
02:49 thiago and don't make any further changes, because you can't commit anything
02:49 lotus I want 99% of the changes in the merge though
02:49 MattMaker joined #git
02:50 thiago lotus: ok
02:50 lotus I guess I need to resolve the conflict is what you're saying.
02:50 thiago anything below 100% resolved can't be committed
02:51 lotus makes sense.  And leaving a file out doesn't really work because it's part of the HEAD of that branch and if the HEAD was merged that file would be assumed merged also
02:51 thiago correct
02:52 * lotus gets to resolving the merge conflict...
02:53 danimal2 joined #git
02:56 f3r70rr35f joined #git
02:56 arescorpio joined #git
02:57 cqi joined #git
02:57 dmj` joined #git
02:58 Dougie187 left #git
02:58 boombatower joined #git
03:01 sanketdg joined #git
03:01 chachasmooth joined #git
03:02 stantonk joined #git
03:02 romerocesar joined #git
03:05 MattMaker joined #git
03:05 Sasazuka joined #git
03:06 Literphor joined #git
03:06 auzty joined #git
03:07 chrisshattuck joined #git
03:07 xall joined #git
03:11 shgysk8zer0_ joined #git
03:11 MattMaker joined #git
03:14 stantonk joined #git
03:16 MattMaker joined #git
03:16 kbs joined #git
03:17 hahuang61 joined #git
03:17 d4rklit3 joined #git
03:19 danimal2 joined #git
03:20 theskillwithin joined #git
03:22 romerocesar joined #git
03:25 pioiytr__ joined #git
03:27 treaki__ joined #git
03:29 Guest39441 joined #git
03:32 stantonk joined #git
03:32 romerocesar joined #git
03:33 MattMaker joined #git
03:40 stoopkid joined #git
03:42 Garen joined #git
03:48 mehola joined #git
03:48 zacts joined #git
03:48 danimal2 joined #git
03:49 rominronin joined #git
03:50 howdoi joined #git
03:52 Masterphi joined #git
03:52 tristanp joined #git
03:52 hk238 joined #git
03:52 Literphor joined #git
03:55 MattMaker joined #git
04:03 romerocesar joined #git
04:03 armyriad joined #git
04:06 d4rklit3 joined #git
04:10 rscata joined #git
04:12 stantonk joined #git
04:14 drewo joined #git
04:18 rchavik joined #git
04:19 sctskw joined #git
04:19 govg joined #git
04:21 nilg joined #git
04:22 romerocesar joined #git
04:23 Rodya_ joined #git
04:24 stonerfish joined #git
04:25 aidalgol In gitk, what is the difference between "Reload" and "Update"?
04:25 stantonk joined #git
04:26 JanC_ joined #git
04:27 mehola joined #git
04:29 Literphor joined #git
04:30 rchavik_ joined #git
04:34 drewo joined #git
04:34 theskillwithin joined #git
04:34 lordjancso joined #git
04:36 chrisshattuck joined #git
04:37 MattMaker joined #git
04:38 chrisshattuck joined #git
04:39 rgrinberg joined #git
04:40 Cabanossi joined #git
04:42 chrisshattuck joined #git
04:44 chrisshattuck joined #git
04:47 FuzzySockets joined #git
04:50 rominronin joined #git
04:50 xall joined #git
04:52 danimal2 joined #git
04:53 romerocesar joined #git
04:57 freimatz joined #git
05:01 mariuscc joined #git
05:03 chrisshattuck joined #git
05:03 romerocesar joined #git
05:04 Duder9999 joined #git
05:06 phanimahesh joined #git
05:11 tristanp joined #git
05:16 cd-rum joined #git
05:16 phanimahesh joined #git
05:19 cdown joined #git
05:21 chrisshattuck joined #git
05:23 romerocesar joined #git
05:27 ThomasLocke joined #git
05:28 f3r70rr36f joined #git
05:29 BrianBlaze420 joined #git
05:29 xall joined #git
05:33 Dragon535 joined #git
05:34 milki no clue!
05:34 milki what does it look like?
05:34 danimal2 joined #git
05:35 drewo joined #git
05:40 OnBrokenWings joined #git
05:41 jungsubk joined #git
05:42 emPi joined #git
05:43 danimal2 joined #git
05:45 jstimm_ joined #git
05:46 drewo joined #git
05:46 qt-x joined #git
05:46 zeroed joined #git
05:46 jceb joined #git
05:49 stantonk joined #git
05:50 Alienpruts joined #git
05:51 rominronin joined #git
05:55 tobiasBora joined #git
05:59 drewo joined #git
06:00 xlegoman joined #git
06:01 leeN joined #git
06:03 LeBlaaanc joined #git
06:04 pioiytr joined #git
06:05 jceb joined #git
06:06 Darren_ joined #git
06:06 romerocesar_ joined #git
06:09 DieguezZ joined #git
06:09 romerocesar joined #git
06:11 circ-user-foXmx joined #git
06:12 b_rare joined #git
06:14 mehola joined #git
06:17 zeroed joined #git
06:17 zeroed joined #git
06:17 romerocesar joined #git
06:21 mkoskar_ joined #git
06:21 mwhooker joined #git
06:23 hendry joined #git
06:23 zlg joined #git
06:24 hendry hi i am on a branch "foobar" and I want to push that to "aremote" to master. What's the syntax for that?
06:24 kimon joined #git
06:24 mwhooker joined #git
06:24 grawity git push aremote foobar:master
06:24 pioiytr_ joined #git
06:24 hendry grawity: thank you
06:25 zlg I have two files with changes that I want to commit, but only after a rebase. I need to commit them one at a time. git-stash seems like the right solution, but don't I need to put both into the index first? If I do that, then 'git stash pop' after my pull, can I remove the second file from the index and commit (with un-added changes) like expected?
06:25 phanimahesh joined #git
06:25 grawity stash, rebase, reset, re-add
06:26 grawity both `git add` and `git reset` take filenames, and even have -p
06:26 grawity alternatively, stash --include-untracked
06:26 freimatz joined #git
06:26 zlg grawity: perfect, thanks for the reassurance.
06:29 mkoskar_ joined #git
06:29 hendry grawity: how do i specify the local branch to push in .git/config
06:35 ynanm joined #git
06:36 muelli joined #git
06:37 Khisanth joined #git
06:37 Alienpruts joined #git
06:38 grawity hendry: I think `git branch -u aremote/master` also applies to pushing
06:38 grawity hendry: as long as push.default is 'upstream'
06:39 grawity see also `man git-config` for branch.*.{remote,pushRemote,merge}
06:39 gitinfo the git-config manpage is available at http://jk.gs/git-config.html
06:39 grawity also remote.pushDefault
06:41 tristanp joined #git
06:43 JeroenT joined #git
06:43 usm4n joined #git
06:45 drewo joined #git
06:49 infinite_ joined #git
06:50 romerocesar_ joined #git
06:52 rominronin joined #git
06:52 CausaMortis joined #git
06:54 romerocesar joined #git
06:57 tobiasBora joined #git
06:59 qt-x1 joined #git
07:00 Repox joined #git
07:01 romerocesar_ joined #git
07:04 infinite_ joined #git
07:04 ahmed_elgabri joined #git
07:04 djb-irc joined #git
07:04 _Patrick_ joined #git
07:06 deBase joined #git
07:06 usm4n joined #git
07:08 gfixler joined #git
07:09 JeroenT joined #git
07:12 lindenle joined #git
07:13 k3rn31 joined #git
07:13 TomyWork joined #git
07:14 usm4n_ joined #git
07:15 Hounddog joined #git
07:16 _Patrick_ joined #git
07:16 dsrtrck joined #git
07:17 Juri_Agarin joined #git
07:17 Fijit joined #git
07:17 encod3 joined #git
07:21 hahuang61 joined #git
07:24 JeroenT joined #git
07:27 antranigv how can I set the `project name` of a git repo? :) right now it's showing `.git` :)
07:27 thekingofbandit joined #git
07:28 Balliad joined #git
07:30 OnBrokenWings joined #git
07:31 usm4n_ left #git
07:34 aswen joined #git
07:38 lindenle joined #git
07:39 grawity you can try writing it into .git/description
07:39 tobiasBora joined #git
07:40 codebam my last commit was a merge, I messed it up (missed some stuff), fixed that, how do I rewrite the commit with the things I fixed inside?
07:41 codebam or well I don't want to change the commit message, just want to add the changes
07:41 rominronin joined #git
07:42 grawity git commit --amend
07:42 grawity you can add -C/--reuse-message HEAD
07:43 codebam oh cool. thank you
07:43 jceb joined #git
07:44 djb-irc joined #git
07:46 Macaveli joined #git
07:46 devbug joined #git
07:46 mikecmpbll joined #git
07:47 Dragon535 left #git
07:48 cyan__ joined #git
07:49 romerocesar joined #git
07:49 kupad joined #git
07:49 drewo joined #git
07:49 codebam if I have a file like a png file, that isn't code. how can I merge those conflicts?
07:50 multi_io joined #git
07:50 grawity you can choose one version with `git checkout --ours/--theirs <file>`
07:51 theskillwithin joined #git
07:52 codebam oh okay thats cool, thanks
07:53 lugzer joined #git
07:55 kupad left #git
07:55 grawity ah I suppose if you want to keep both there's `git show :2:data/foo.png > data/foo-ours.png; git show :3:data/foo.png > data/foo-theirs.png`
07:57 Silenced joined #git
07:58 Silenced joined #git
08:01 mischat joined #git
08:02 Repox_ joined #git
08:04 geoff-codes joined #git
08:04 lindenle joined #git
08:05 Repox_ left #git
08:05 mikecmpbll joined #git
08:05 Repox joined #git
08:06 bongjovi joined #git
08:07 shirty joined #git
08:07 pioiytr joined #git
08:10 romerocesar joined #git
08:11 kurkale6ka joined #git
08:12 muelli joined #git
08:14 stoopkid joined #git
08:14 stoopkid joined #git
08:14 King_DuckZ joined #git
08:16 aswen joined #git
08:17 King_DuckZ hi, is there a way to list the remote url of each submodule without initializing them first?
08:18 theskillwithin joined #git
08:19 Beams joined #git
08:20 Andrew_K joined #git
08:21 Duikb00t joined #git
08:21 Duikb00t o/
08:21 CausaMortis-Work joined #git
08:21 hahuang61 joined #git
08:22 apotry joined #git
08:27 theskillwithin joined #git
08:28 jceb joined #git
08:30 romerocesar joined #git
08:30 f3r70rr35f joined #git
08:32 Jellyg00se joined #git
08:32 t0by joined #git
08:34 tristanp joined #git
08:34 clmsy joined #git
08:35 kab joined #git
08:36 mikecmpb_ joined #git
08:37 kedare joined #git
08:38 DieguezZ joined #git
08:39 codebam how do I pull in one file from a remote tracking branch?
08:40 OnkelTem joined #git
08:41 theskillwithin joined #git
08:41 OnkelTem Hi all. Need urgent help. I wanted to join two last commits and for this I launched: git rebase -i HEAD~2
08:41 codebam never mind, got it
08:42 OnkelTem but for some strange reason it opened editor with a bunch of commits to squash
08:42 OnkelTem Now I don't konw what to do! If I exit the editor it would probably fuck up my repository
08:42 OnkelTem how to cancel the action? kill the process?
08:43 drewo joined #git
08:46 PCatinean joined #git
08:46 kab Just comment lines
08:47 kab it'll do nothing
08:48 alcohol OnkelTem: which editor?
08:48 OnkelTem alcohol: I have mcedit here
08:49 alcohol world editor for minecraft? o_O
08:49 OnkelTem kab: oh, thanks, I'll try
08:49 OnkelTem alcohol: lol
08:49 alcohol well thats the first google result.. http://www.mcedit.net/
08:49 dreiss joined #git
08:49 OnkelTem alcohol: Midnight Commander
08:50 alcohol ah
08:51 OnkelTem kab: many thanks it worked!
08:51 GenZai joined #git
08:52 kab <OnkelTem> now analyse why it has happened and tell us )
08:55 FunkyAss joined #git
08:56 Limes_ joined #git
08:58 Masterphi joined #git
08:58 kab joined #git
09:01 alcohol kab: it happened cause he did an interactive rebase
09:01 alcohol kab: i think he just didnt read up on how it works
09:01 kab his command should show onle two commits for rebase, not a bunch as he has
09:02 kab i use that command every day though
09:02 kab maybe his head was not he expected
09:03 tvw joined #git
09:03 alcohol OnkelTem: what did you see exactly?
09:03 achadwick joined #git
09:03 alcohol OnkelTem: if you saw 2 commits prefixed with "pick", then that was exactly what you should have expected
09:04 kab yep it should be like that in normal state )
09:04 alcohol OnkelTem: and you probably would have wanted to change one of those to fixup or squash
09:06 pioiytr_ joined #git
09:06 ikmaak joined #git
09:10 theskillwithin joined #git
09:10 kab joined #git
09:11 chll_ joined #git
09:13 kab joined #git
09:13 sdothum joined #git
09:14 jceb joined #git
09:14 ikmaak joined #git
09:15 nnyk_ joined #git
09:17 ncd|leeN joined #git
09:17 troulouliou_div2 joined #git
09:19 JeroenT joined #git
09:22 pcre_ joined #git
09:22 notebox joined #git
09:22 lindenle joined #git
09:23 King_DuckZ I need to get the default remote exactly as asked in this question http://stackoverflow.com/questions/34099451/get-default-remote-push-and-default-remote-branch is there anybody with an answer for it?
09:23 Beams joined #git
09:23 King_DuckZ I would also like to avoid any manual splitting, as the guy in that thread asked
09:23 hahuang61 joined #git
09:23 notebox joined #git
09:26 durham joined #git
09:27 armyriad joined #git
09:28 johnny56_ joined #git
09:29 solf1re joined #git
09:29 solf1re joined #git
09:30 solf1re joined #git
09:31 solf1re joined #git
09:32 solf1re joined #git
09:32 liamonade joined #git
09:32 theskillwithin joined #git
09:33 solf1re joined #git
09:33 solf1re joined #git
09:38 drewo joined #git
09:41 white_gecko joined #git
09:43 des_consolado Hey if I use a hard link from a directory within my repo to another directory within my repo, will git be okay with that or no?
09:43 white_gecko Hello, I have two questions:
09:43 white_gecko Is there any scientific literature about git, especially something like a formal mathematical model about commits, merges, diffs, rebase, … ?
09:43 white_gecko In which cases is it not enought for a three way merge to take the latest common acestor of two commits? When do you actually need recursive three way merge?
09:44 j416 des_consolado: git will treat them as any other file
09:44 des_consolado Basically I'm building a repo for a docker build of one of our projects, and there are many dependencies, for this particular dependency there is another repo of ours, so I include the repo as a submodule and it's part of this repo, but for the docker container itself when it is ran, it mounts a dir in the repo called lib/, so I'd like to use a hardlink to have a file in a dir in lib be linked to a
09:44 des_consolado file in the submodule
09:45 nuc joined #git
09:45 des_consolado but if another user clones this repo and updates the file that is the hard link, will that also update the file in the sub repository?
09:45 j416 no
09:45 j416 des_consolado: ^
09:46 des_consolado (i.e. will the hard link be maintained through push/pull to and from remote, is git aware of inodes and all that?)
09:46 des_consolado ah, okay
09:46 j416 des_consolado: git treats it as a regular file, nothing special, it does not care
09:46 phanimahesh joined #git
09:46 j416 des_consolado: it does do symlinks, though
09:51 des_consolado yeah docker doesn't do symlinks :P
09:51 des_consolado it's cool though, just done some extra docker mounts
09:52 stantonk joined #git
09:54 mikecmpbll joined #git
10:01 LiohAu If I do a "git merge develop -X theirs", does it takes "develop" version? or current branch version? develop right?
10:02 canton7 yes
10:02 LiohAu ok, so why does git tells me there's nothing to commit after that?
10:02 LiohAu I should have to commit the changes coming from develop, no?
10:02 canton7 because the merge completed, created the merge commit, and you haven't made any changes since the merge?
10:02 LiohAu no, only the merge
10:03 canton7 ok, so the merge completed. why do you expect there to be uncommitted changes?
10:03 canton7 'git log' will show it
10:03 LiohAu so basically I have a PR open with 1 changed file. That PR cannot automaticlaly merged because someone changed the same file and pushed to develop
10:03 LiohAu I only want to make that PR automatically mergeable agian
10:03 LiohAu again*
10:04 canton7 ok, so you'll need to merge develop into your PR branch
10:04 canton7 you probably don't want to use -X theirs - that will override your changes in the case that there's a conflict
10:04 LiohAu ok, so what I did with a "git merge develop -X theirs" + "git checkout --HEAD my_change_file"
10:04 cdown joined #git
10:05 LiohAu but I don't have any other change to make, so how do I push that new "state"
10:05 canton7 why the checkout? that will just remove an uncommitted changes to my_change_file
10:05 canton7 *remove any
10:05 canton7 ... but since there shouldn't be any uncommitted changes anyway, it won't do anything
10:05 LiohAu because that file I want to keep "ours" version
10:06 LiohAu I read on S.O. that it is the same as "git checkout --ours -- <filename>"
10:06 canton7 1) you don't combine two sets of changes by completely replacing what the other person did, and 2) like I said, that won't have done what you wanted it to do
10:06 canton7 that checkout only does something if you're stopped in the middle of a conflicted merge
10:06 canton7 your merge won't have conflicted, because -X theirs automatically resolves all conflicts in favour of them
10:07 LiohAu Well in this case I want to replace its changes for that file, because he changed things without knowing that the whole file was being refactored..
10:07 LiohAu so, then how do I get all the new things from develop, and keep my changed file?
10:07 canton7 then you'll remove all of the changes he made, don't forget
10:08 canton7 undo your merge, do it again *without* -X theirs, when the merge stops because of a conflict, use 'git checkout --ours <filename>', then 'git commit' to complete the merge
10:08 canton7 ... then you'll have to re-do the changes he made by hand
10:09 cyan__ joined #git
10:09 LiohAu "then you'll remove all of the changes he made" => On that specific file yes, but I want to keep all the other changes he did.
10:09 Meow-J joined #git
10:09 LiohAu what? so if he changed 100 files, and we have a conflict on 1 file, I have to redo all the 100 changes?
10:09 canton7 yeah, I was refererring to that particular file
10:09 canton7 no
10:09 LeMike joined #git
10:09 canton7 just on that one file
10:10 LiohAu oh ok, but as I said i don't get anything from him on that file
10:10 ArchNoob joined #git
10:10 canton7 you're saying he didn't change '<filename>'?
10:11 LiohAu he did, but he should have undo when I told him that file was being deeply changed/refactored. And he forgot to undo, and pushed that file with the other changes..
10:11 ikmaak joined #git
10:11 canton7 ok
10:11 canton7 the commands I gave above still apply then
10:12 LiohAu ok so right now, "git reset --hard ORIG_HEAD" to cancel my "invalid merge" right?
10:12 LiohAu and your commands
10:12 pcre joined #git
10:12 canton7 yeah, unless you've done something since to reset ORIG_HEAD
10:12 canton7 do you understand why what you were trying to do was failing?
10:13 LiohAu I think that the file I was trying to checkout with --ours, was not in a conflict state, so the checkout was doing nothing, right?
10:14 noidea joined #git
10:14 Beams_ joined #git
10:14 canton7 it's not just the file that wasn't in a conflict state: nothing was. If you use -X theirs, the merge will not conflict - the point of -X theirs is to automatically resolve conflicts. So nothing was conflicted. You just run a command to replace the contents of the file with what was in the previous commit (i.e. replace the result of the merge with the result of the merge)
10:15 LiohAu actually I think I have a problem in the way conflicts are handled in git.
10:15 canton7 I doubt that, but go on
10:17 p4trix joined #git
10:18 mika joined #git
10:19 King_DuckZ I want to parse the output of git submodule status to get the current state of each submodule (the +/- first char), is there a more scripting-friendly command to achieve the same? initialized submodules will have a trailing (heads/maste) which is causing extra work
10:20 CrazyBonz joined #git
10:20 indieross joined #git
10:21 romerocesar joined #git
10:21 thalleralexander joined #git
10:24 thalleralexander joined #git
10:24 hahuang61 joined #git
10:25 pcre joined #git
10:25 nellicus_ joined #git
10:26 LiohAu canton7, sorry got a call. So, I was saying that I may have misunderstood something, because when I merge without using ours/theirs what I see in the list of changes always makes me thing that my branch is taken in priority over the develop one.
10:26 nnyk_ joined #git
10:26 LiohAu this is why I wanted to use the -theirs
10:28 canton7 a merge combines your changes with theirs - neither takes priority
10:29 canton7 if you reach a case where both yours and theirs make the same change, so there's a question about which has priority, git creates a conflict
10:30 msonntag joined #git
10:31 boombatower joined #git
10:32 drewo joined #git
10:32 waveclaw joined #git
10:32 ikmaak joined #git
10:33 LiohAu oh fuck
10:33 netj joined #git
10:33 LiohAu git made an automatic merge on a file which is automatically generated from the one I told him to take my version
10:34 canton7 you probably shouldn't be tracking generated files
10:34 canton7 ... but anyway, you can fix that file up, and git commit --amend to change the outcome of the merge
10:34 fahadash joined #git
10:34 LiohAu yeah I know, we are going to generate them in the build process soon :(
10:35 canton7 ... but be sure to never edit a merge commit so that it contains changes which didn't come from either parent - those are so-called evil merges, and it's really hard to track down where the change was introduced
10:35 mehola joined #git
10:36 pcre___ joined #git
10:36 rchavik joined #git
10:36 LiohAu I don't even know how to fix my mistake
10:36 canton7 1) run the script which generates the build output, 2) 'git add generatedfile', 3) 'git commit --amend'
10:36 LiohAu I used the "git checkout --ours -- " as you said
10:37 canton7 wait, stop
10:37 canton7 are you still in the middle of a conflicted merge
10:37 canton7 wait, you ran 'git checkout --ours --' and not 'git checkout --ours thefile'?
10:37 LiohAu git checkout --ours -- thefile
10:37 LiohAu no i'm not in the middle of the conflicted merge anymore
10:37 canton7 you scared me there
10:38 canton7 ok, then the instructions I gave still hold
10:38 johnmilton joined #git
10:40 johnmilton joined #git
10:41 JeroenT joined #git
10:42 pcre_ joined #git
10:42 LiohAu canton7: https://gist.github.com/LiohAu/4dff682428a92538c857532b85d8a7ce
10:42 johnmilton joined #git
10:42 LiohAu this is the log of what I did
10:43 * canton7 reads
10:43 xall joined #git
10:43 canton7 About_Page_Mobile_64256.feature.cs is the file which you both modified?
10:43 LiohAu the .feature file is the file we both modified. The .feature.cs is an automatically generated one from the associated .feature
10:44 canton7 ok, but you ran 'git checkout --ours' on the generated file?
10:44 johnmilton joined #git
10:44 canton7 hmm, About_Page_Mobile_64256.feature didn't conflict
10:44 canton7 so you didn't edit the same lines in it
10:44 LiohAu that's what is weird, you see, when I pulled with "git pull origin feature_file-about", it was like if the merge I did before was still there
10:45 canton7 I don't know whether it was...
10:45 tvw joined #git
10:45 canton7 output of !lol ?
10:45 gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all
10:46 LiohAu updated gist : https://gist.github.com/LiohAu/4dff682428a92538c857532b85d8a7ce
10:47 johnmilton joined #git
10:47 johnmilton joined #git
10:47 canton7 ok right, so you did the merge twice by the looks of it
10:47 LiohAu how do you see it?
10:47 LiohAu oh the two "comments" saying the same thing
10:48 LiohAu what does a star and a line means ?
10:48 yqt joined #git
10:49 canton7 stars and commits. the line next to a star is the "commit_hash (branch name) commit message"
10:49 Codfection joined #git
10:49 mrkat joined #git
10:49 canton7 lines join commits together - they show which commits are the parents and children of which other commits
10:49 canton7 time begins at the bottom, with the newer stuff on top
10:49 canton7 so af0387d merges together 6cc1fbf and 167387a
10:50 canton7 and 9dbcb99 does the same thing
10:50 vls-xy joined #git
10:50 canton7 so it looks like you did a merge, pushed it, then undid it locally and did it again
10:50 canton7 ... which is fine
10:50 canton7 but merging in origin/feature_file-about (which contains your first attempt to do the merge, which you later undid) would be a mistake
10:50 Impaloo joined #git
10:51 canton7 ... being able to visualise git history is the first step to learning how to use git. The ascii output is useful for pasting over IRC, but is slightly cryptic to interpret. There are much better graphicaly tools: gitk is often distributed with git, or there are others
10:51 JeroenT joined #git
10:51 LiohAu i'm on osx :(
10:51 johnmilton joined #git
10:52 romerocesar joined #git
10:52 canton7 I've got gitk on osx - can't remember whether it came pre-installed, with the homebrew git version, or whether it was in a separate homebrew package...
10:52 canton7 there are other visualisers for mac as well
10:53 vls-xy perhaps gitx suits your fancy: http://rowanj.github.io/gitx/
10:53 MrWoohoo joined #git
10:53 tristanp joined #git
10:53 LiohAu Just installed the homebrew version, and its comes with it (default apple git does not have it)
10:53 CheckDavid joined #git
10:54 acetakwas joined #git
10:54 canton7 cool
10:54 LiohAu yeah gitk looks a little bit like 90's software :P
10:54 canton7 'gitk --all &' in a git repo is a good way to launch it
10:55 canton7 yeah, it's not pretty on osx - it's tcl/tk iirc, which gets rendered pretty badly
10:55 canton7 it's better on windows, or even better on linux :P
10:55 bremner it's not _pretty_ anywhere, but mac people seem to care more ;)
10:55 canton7 it *is* worse-looking on mac though
10:55 bremner haha
10:55 bremner justice!
10:55 canton7 ><
10:56 cdown joined #git
10:56 LiohAu well maybe because we are used to see pretty things :P
10:56 Beams joined #git
10:56 LiohAu I would say that linux is for working stuff, windows is for beautilfull stuff, and osx is for having both :P
10:59 theskillwithin joined #git
11:01 pcre__ joined #git
11:01 CussBot joined #git
11:02 rnsanchez joined #git
11:04 dimi1947 joined #git
11:04 Impaloo joined #git
11:09 LiohAu canton7: I feel like if I were a variable in an infinite loop :'(
11:10 LiohAu tried to push, was rejected because : hint: Updates were rejected because the tip of your current branch is behind...
11:10 canton7 yep, because you changed a commit which you'd already pushed. you !rewrote public history
11:10 LiohAu so I pulled, and the same conflict as the one I solved just before
11:10 canton7 nooo don't pull here!
11:10 canton7 bleh, !rewrite
11:10 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
11:11 LiohAu :'(
11:11 canton7 normally, force-pushing is a bad idea. However it's very useful with PR branches: noone else is working from them, and github knows how to handle them, in order to change which commits are part of the PR
11:11 canton7 so in this case you want to fix up your local branch so it has just the right merge in, the force-push it
11:12 theskillwithin joined #git
11:12 LiohAu so how do I cancel my pull ? without an rm -rf :P
11:12 canton7 let's see that !lol output again
11:12 gitinfo A nifty view of branches, tags, and other refs: git log --oneline --graph --decorate --all
11:12 canton7 we'll find a suitable commit to reset back to, and reset to it
11:13 LiohAu https://gist.github.com/LiohAu/4dff682428a92538c857532b85d8a7ce
11:13 LiohAu you have both the previous log, and the new one
11:14 canton7 right, so we can see what the merge commit af0387d was merged into the merge commit 9dbcb99
11:14 canton7 wait, no it doesn't
11:14 canton7 ignore that
11:14 ArchNoob joined #git
11:15 canton7 right, so merge merged together 167387a and 6cc1fbf to make 9dbcb99, then edited that to make a7a44e4?
11:15 LiohAu basically I want to keep 167387a
11:15 LiohAu yeah I want to keep 167387a  + 6cc1fbf
11:16 ojdo joined #git
11:16 canton7 so what's wrong with a7a44e4?
11:17 Beams_ joined #git
11:18 LiohAu I don't know, I don't even know how to know what is inside :(
11:18 canton7 I'll reword: is there anything wrong with a7a44e4?
11:18 canton7 'git show' will show you its diff
11:18 canton7 the "inside" of it is currently checked out in your working directory
11:19 acetakwas joined #git
11:19 LiohAu on the remote repo, there is af0387 which is half of what I wanted. The develop content is correctly integrated in the feature_file-about branch, the .feature file content is ok. Only the .feature.cs file is not.
11:19 TomyWork Does the "tag" in "git pull $url tag $tag" make any sense there?
11:20 canton7 Tomno
11:20 canton7 *TomyWork, no
11:20 TomyWork Jenkins writes that and I think it's an error on its behalf.
11:20 TomyWork git pull failed
11:20 TomyWork no further info, unfortunately
11:20 canton7 LiohAu, shall we start over? Get this right first time, no confusion?
11:21 LiohAu git show of a7a44e4 looks ok (If what is green is what is in this commit, then its what I want)
11:21 canton7 what is green is what was introduced in that commit
11:22 HoierM joined #git
11:22 romerocesar joined #git
11:22 canton7 anyway, if a7a44e4 contains what you want to keep, then you can replace what's on origin with it, by using 'git push -f origin feature_file-about'
11:22 rahtgaz joined #git
11:22 canton7 note that force-pushing is usually a VERY BAD IDEA. But it's fine in some cases, and PR branches is one of those
11:23 white_gecko joined #git
11:23 notebox joined #git
11:24 LiohAu ok, but right now I can't push
11:24 LiohAu because my repo is in a conflicted state
11:24 LiohAu (I don't know why git log shows nothing about this)
11:24 canton7 wait, you're in the middle of a merge commit? ok
11:24 canton7 abort that: 'git merge --abort'
11:24 LiohAu y
11:25 LiohAu ok
11:25 xpk joined #git
11:25 LiohAu what happens when I do that? ok, it cancel the merge, but what happens to the pull that raised the conflict ?
11:25 hahuang61 joined #git
11:26 white_gecko Is there any scientific literature about git?
11:26 drewo joined #git
11:26 canton7 LiohAu, hmm? do you know that 'pull' is just shorthand for 'fetch + merge'?
11:26 LiohAu white_gecko: is this scientific : https://git-scm.com/doc ?
11:26 canton7 so 'git pull origin master' is shorthand for 'git fetch origin; git merge origin/masteer'
11:27 LiohAu canton7: no, but know I do :P
11:27 canton7 so the 'git merge --abort' aborts the merge portion, but the fetch portion is left untouched
11:27 NeonSynth joined #git
11:28 LiohAu ok, so you tell me to push -f
11:28 canton7 yep
11:28 LiohAu but git status tells there's no pending commit
11:28 canton7 pastebin?
11:28 LiohAu $ git status
11:28 LiohAu On branch feature_file-about nothing to commit, working directory clean
11:29 canton7 that means you have no uncommitted changes
11:29 canton7 that means that a 'git commit' would be pointless
11:29 canton7 but you still have local *commits* which origin doesn't have
11:29 canton7 so you can push stuff
11:29 rchavik joined #git
11:29 LiohAu shouldn't git says something like "my branch is 1 commit ahead" or something like this?
11:29 canton7 it'll probably say 2 ahead / 1 behind, if you've got tracking set up
11:29 hk238 joined #git
11:30 canton7 but you probably don't have tracking set up for that branch. Don't worry about it
11:30 white_gecko LiohAu: It is a technical documentation so for computerscience, yes partially.
11:30 osse white_gecko: !bottomup
11:30 gitinfo white_gecko: '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)
11:30 white_gecko but i was especially looking for something like a formal mathematical model about commits, merges, diffs, rebase, … ?
11:31 jatin30 joined #git
11:31 david joined #git
11:31 LiohAu canton7: https://gist.github.com/LiohAu/4dff682428a92538c857532b85d8a7ce
11:31 canton7 LiohAu, yes?
11:32 LiohAu the feature_file-about has no arrow like the HEAD one, this means there's no tracking setup ?
11:32 cdown joined #git
11:32 canton7 yes
11:32 canton7 'git branch --set-upstream-to origin/feature_file-about'
11:32 canton7 'git branch --set-upstream-to=origin/feature_file-about'
11:32 david007 joined #git
11:32 LiohAu I thought tracking was automatically setup when you make a "git checkout -b branch_name" and then "git push origin branch_name"
11:33 canton7 LiohAu, it's automatically set by 'git checkout -b branch_name origin/branch_name'
11:33 canton7 'push' doesn't set it unless you pass it -u
11:33 jatin30 when I cherry-pick and get merge conflicts I resolve them manually using mergetool, meld. I do not read on the whole code I just add anything which is extra and kindof resolve the merge conflicts randomly which just does not seem right. Any correct way of doing it? Should I use git pull --rebase ?
11:34 LiohAu so I should change my habits when I create a feature branch and use "git checkout -b branch_name origin/branch_name" instead of "git checkout -b branch_name" ?
11:34 white_gecko osse: I need a password for the newartisans, but is the ikke.info link the official one?
11:34 canton7 LiohAu, no! they do completely different things
11:34 canton7 'git checkout -b branch_name' creates a new local branch, branch_name, based off your currently-checked-out branch
11:34 canton7 'git checkout -b branch_name origin/branch_name' creates a new local branch, branch_name, based off the remote branch origin/branch_name
11:35 canton7 you use the former to create a new branch, and the latter to start working on a branch which someone else created
11:35 osse white_gecko: I don't know. _ikke_ is the guy hosting it.
11:35 osse but it seems to be the exact same document
11:35 PioneerAxon joined #git
11:36 _ikke_ I'
11:36 _ikke_ I'm mirroring it because it was down some time ago
11:37 mehola joined #git
11:37 _ikke_ The ftp one works for me though
11:38 osse it's not a dead link, but I have to log in
11:39 _ikke_ Hmm, I don't get a login prompt
11:40 jatin30 why does this not work https://bpaste.net/show/82bdda582874 ?
11:41 Beams joined #git
11:42 _ikke_ You have a conflict
11:42 osse jatin30: the patch doesn't apply cleanly
11:42 _ikke_ You need to fix the conflict
11:42 white_gecko is there a repo with the sources of thei pdf?
11:42 osse unfortunate, but not unexpected
11:43 jatin30 osse, what should I do to fix this?
11:43 osse jatin30: !eek
11:43 gitinfo jatin30: [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice! A nice video explaining merge conflicts: https://www.youtube.com/watch?v=zz7NuSCH6II
11:43 flaguy48 left #git
11:43 osse 'git status' and 'git diff' are good starting points also
11:43 raindev joined #git
11:44 MrMojito joined #git
11:45 jatin30 osse, there are good chances that I might have randomly resolved some previous merge conflicts by no proper judgement as to which part of file is more relevant. I just made the three copies exactly similar using mergetool and finished the cherry-pick. what is the worst thing that could happen?
11:46 mdw joined #git
11:46 Noldorin joined #git
11:47 osse jatin30: the worst that could happen is that the source code in the mentioned file won't compile, or will do something else than intended
11:47 mdw joined #git
11:50 jatin30 joined #git
11:51 jatin30 sorry my system hung. osse, could you please copy your ans to my last question please, if you did ans.
11:51 jatin30 !eek
11:51 gitinfo [!eekaconflict] Merge conflicts are a natural part of collaboration. When facing one, *don't panic*. Read "How to resolve conflicts" in man git-merge and http://git-scm.com/book/ch3-2.html#Basic-Merge-Conflicts then carefully go through the conflicts. Picking one side verbatim is not always the right choice! A nice video explaining merge conflicts: https://www.youtube.com/watch?v=zz7NuSCH6II
11:51 osse jatin30: the worst that could happen is that the source code in the mentioned file won't compile, or will do something else than intended
11:51 cdown joined #git
11:51 jatin30 osse, I was afraid of that! any way I could stop that?
11:52 osse jatin30: either fix the conflict or abort the cherry-pick
11:52 JeroenT joined #git
11:52 stratos joined #git
11:53 white_gecko In which cases is it not enought for a three way merge to take the latest common acestor of two commits? When do you actually need recursive three way merge?
11:53 jatin30 osse, conflicts are fixed. cherry picking is done
11:54 jatin30 let me test it on my phone and see what happens
11:55 mahendra_ joined #git
11:56 bruce_lee joined #git
11:56 bruce_lee joined #git
11:56 jatin30 but after every cherry-pick I do a repo sync before the next cherry-pick and it runs fine. as repo sync runs fine I am guessing Its compiling properly right?
11:57 rchavik joined #git
11:57 osse whether it compiles properly only javac can tell you
11:57 osse not git, and certainly not me
11:58 jatin30 oh yes
11:58 jatin30 that was so lame
11:58 jatin30 ok thanks
11:58 mahendra_ left #git
11:58 osse i don't know what a "repo sync" is, but why are you doing all these cehrry picks?
11:59 synthroid joined #git
12:00 jatin30 we are building on cyanogenmod for making an OS
12:00 jatin30 so cherry picking basic features for that
12:00 synthroi_ joined #git
12:01 notebox joined #git
12:01 flaguy48 joined #git
12:01 tvw joined #git
12:02 DieguezZ joined #git
12:03 tomog999_ joined #git
12:06 pioiytr joined #git
12:06 wspider I have two branches with folder 'foo', I deleted that folder from branch A because I want both to have the foo content of B, so I deleted foo from A and tried to  merge with B
12:07 wspider mergin A with B or B with A, the result doesn't have the foo folder so i 'git reflog' and reset --hard according and am up to trying new things until I get what I wanted
12:07 liquidjamm joined #git
12:07 wspider any hint ?
12:07 LiohAu wspider: canton7 will help you :P
12:09 canton7 hey, I'm eating lunch: give me a break :)
12:09 bremner /kick canton7
12:09 bremner as requested ;)
12:09 wspider LiohAu: haha :>
12:10 LiohAu :D
12:10 LiohAu i'm gonna eat too
12:10 LiohAu canton7: are you in CET area ?
12:10 canton7 ish - GMT
12:10 TestUser joined #git
12:11 LiohAu hm I see you're on c# and iphonedev channels, are you a xamarin user ?
12:12 LiohAu (or a polyglot :P)
12:12 NeverDie joined #git
12:14 NeverDie joined #git
12:14 wspider halp
12:14 wspider xd
12:15 NeverDie joined #git
12:15 maaarghk joined #git
12:16 canton7 LiohAu, I work at a consultancy, so a bit of everything
12:17 maaarghk Hello. In 2.9 --allow-unrelated-histories was added as an option to "git merge". The release note below taht says that "git pull" has been taught to pass that option to the underlying git merge. However, git rebase does not appear to have been taught the same thing. Is there a way to pass that option from rebase?
12:17 canton7 there's no 'underlying merge' for a rebase
12:17 canton7 rebase and merge are separate things
12:18 maaarghk git rebase uses merge, right? hence -s and -X options being passed down.
12:18 mdw_ joined #git
12:19 qknight left #git
12:19 rwb joined #git
12:19 canton7 sort of. git rebase takes a set of commits, and one-by-one cherry-picks them onto a branch (well it's not exactly cherry-pick, but close enough)
12:19 canton7 so rebase doesn't take two histories and creates a single merge commit
12:20 canton7 it takes a set of commits, and individually merges each one (effectively) onto a given branch
12:20 drewo joined #git
12:21 Asenar joined #git
12:21 maaarghk yeah, I should have been clearer though and mentioned that I have --preserve-merges as an option to git rebase, which means that merge commits are being created
12:22 maaarghk However they are currently failing with 'refusing to merge unrelated histories' error, and I am trying to work out how to use the git merge --allow-unrelated-histories switch to get past that
12:22 romerocesar joined #git
12:23 wspider LiohAu: I already figured out, thanks !!
12:25 tristanp joined #git
12:25 jeffreylevesque joined #git
12:25 lindenle joined #git
12:25 deadnull joined #git
12:26 hahuang61 joined #git
12:27 toogley joined #git
12:28 aswen joined #git
12:33 romerocesar joined #git
12:35 encod3 joined #git
12:37 mehola joined #git
12:38 ThomasLocke joined #git
12:40 raindev joined #git
12:40 rafalcpp joined #git
12:41 adrianoc joined #git
12:41 AaronF joined #git
12:42 rchavik joined #git
12:43 umbSublime joined #git
12:43 umbSublime joined #git
12:47 adrianoc hi all ... how i can authenticate git clone http://... with ssh key (git clone ssh://)
12:48 BlackMaria joined #git
12:51 liquidjamm joined #git
12:51 rgrinberg joined #git
12:51 spb adrianoc: by using an ssh URL
12:52 adrianoc sbp, ssh is blocked.
12:53 Bombe Then you obviously can not use it for authentication.
12:53 daniel_rose joined #git
12:54 stantonk joined #git
12:54 adrianoc i want to make with a gitHub
12:55 TooLmaN joined #git
12:55 M-shine joined #git
12:55 liquidjamm joined #git
12:56 harish joined #git
12:57 zhbc joined #git
12:57 espindola joined #git
12:57 grawity adrianoc: neither HTTP nor HTTPS uses ssh keys
12:57 grawity adrianoc: you can however use ssh://git@ssh.github.com:443/
12:57 grawity adrianoc: see https://help.github.com/articles/using-ssh-over-the-https-port/
12:58 cuebix-wk joined #git
12:59 Eryn_1983_FL joined #git
12:59 Masterphi joined #git
13:00 adrianoc grawity, it's better.
13:00 grawity ?
13:00 stantonk_ joined #git
13:02 CalimeroTeknik oh, .git folders are ignored by default? I know this is git abuse, but I'm running incremental backups of folders containing git repos using git
13:02 jimi_ joined #git
13:02 CalimeroTeknik using add -Af works even for gitignored files, but not for .git folders
13:02 CalimeroTeknik s/folders/directories/
13:03 mikecmpb_ joined #git
13:05 tjone270_ joined #git
13:05 grawity they're *always* ignored
13:06 adrianoc grawity, i try this way. thanks.
13:06 adrianoc no now, because my server have not http configured.
13:07 rgrinberg joined #git
13:07 adrianoc and how i can't access ssh, just the night to test.
13:08 tchan joined #git
13:08 ExoUNX joined #git
13:08 ExoUNX joined #git
13:10 grawity what server
13:12 crose joined #git
13:13 dimi1947 joined #git
13:14 notebox joined #git
13:14 drewo joined #git
13:15 livingstn joined #git
13:17 usm4n joined #git
13:17 adrianoc grawity, my server, deskfacil.com
13:18 grawity you were talking about github minutes ago
13:18 adrianoc i cant access ssh inside my work.
13:18 Klipp joined #git
13:18 adrianoc grawity, yes, as example.
13:18 ThomasLocke joined #git
13:19 grawity well, if it's not github, then you're out of luck
13:19 DieguezZ joined #git
13:20 adrianoc grawity, thanks, you helpe much.
13:20 adrianoc later i tell you if worked :)
13:20 perlpilot adrianoc: or ... configure your server to handle http(s)
13:21 M-mistake joined #git
13:21 M-meznak joined #git
13:23 ash_workz joined #git
13:24 usm4n joined #git
13:25 adrianoc perlpilot, bu i want using one password.
13:25 grawity configure your server to use TLS certificate auth
13:26 osse King_DuckZ: the more complex parts of that question are probably best asked on the mailing list. As for the superficial parts it seems he answered his own question in that post
13:28 kedare joined #git
13:30 pmst joined #git
13:30 todd_dsm joined #git
13:30 SteffanW joined #git
13:30 daniel_rose joined #git
13:31 g105b joined #git
13:31 marc_v92 joined #git
13:32 circ-user-foXmx joined #git
13:35 g105b I know the answer is going to be "it depends", but I'm struggling to make my own opinion and I'd appreciate it if someone could point me to some recommended reading or their own advice on this. My company automatically deploys the master branch to live, whenever there's a commit. We branch away from master for each issue, or for big changes, introduce
13:35 g105b another branch to branch from for sub-issues. My question is whether we should be rebasing the child branch on the parent branch before each merge? I can't make any sense of why this is regarded as beneficial by some people, but there are many people itching to use rebase over simple merging. Thanks.
13:38 cqi joined #git
13:40 GodGinrai joined #git
13:42 HanClinto joined #git
13:42 rahtgaz joined #git
13:43 osse !merge_vs_rebase > g105b
13:43 gitinfo [!merge_or_rebase] For a detailed discussion of the dis-/advantages of merge and rebase, and when to best use which, see: https://coderwall.com/p/ny1hia -- and here's how Linus explains which one to use: http://www.mail-archive.com/dri-devel@lists.sourceforge.net/msg39091.html
13:43 osse g105b: a major part of it is preference. some like the log being a straight line
13:43 g105b Thanks!
13:45 perlpilot I suspect those people come from svn or some svn-like system
13:45 DieguezZ joined #git
13:46 kbs joined #git
13:46 wodim is it possible to push to all remotes at once?
13:47 madknight wodim, https://stackoverflow.com/questions/5785549/able-to-push-to-all-git-remotes-with-the-one-command
13:47 wodim yeah that's what I was reading. ugly tho
13:47 wodim but ok fair enough
13:47 perlpilot xargs is ugly?
13:48 wodim no, I just thought there would be a git command for it by now--usually stackoverflow answers are outdated
13:48 hwoarang joined #git
13:49 raijin joined #git
13:50 madknight wodim, you can alias it, there is also no command "git squash" but cool aliases
13:57 Gsham joined #git
13:58 Gamecubic joined #git
13:58 Rodya_ joined #git
14:00 GodGinrai wodim: pushing to all remotes at once is an edge-case, which is probably why there isn't a command for it
14:00 wodim not an edge case for me :P
14:00 wodim but ok
14:01 phaleth joined #git
14:02 cdg joined #git
14:03 moriko joined #git
14:03 diegoviola joined #git
14:04 usm4n_ joined #git
14:05 unbalancedparen joined #git
14:05 Darren_ joined #git
14:05 marc_v92 joined #git
14:06 tristanp joined #git
14:08 drewo joined #git
14:09 al-damiri joined #git
14:10 gduine joined #git
14:12 fmcgeough joined #git
14:14 D-Boy joined #git
14:18 dreiss joined #git
14:18 qpdb joined #git
14:19 stantonk joined #git
14:21 Vinnie_win_a joined #git
14:25 dmj` joined #git
14:26 cdown joined #git
14:26 bremner joined #git
14:26 bremner joined #git
14:26 Passe_a_Manteiga joined #git
14:27 l0rdPE joined #git
14:27 hahuang61 joined #git
14:28 mikecmpbll joined #git
14:28 catag87 joined #git
14:29 Asiajey joined #git
14:31 LionsMane joined #git
14:32 circ-user-foXmx joined #git
14:32 hwoarang_ joined #git
14:32 thiago joined #git
14:33 moriko joined #git
14:34 Rodya_ joined #git
14:35 phanimahesh joined #git
14:35 notebox joined #git
14:36 yqt joined #git
14:36 MadHatter42 joined #git
14:38 phanimahesh CalimeroTeknik: Using git to "backup" folders containing git repos is definitely way more than git abuse. please don't do it. :P
14:39 CalimeroTeknik well, I can do tarballs
14:40 CalimeroTeknik but that's going to be even worse since it uses up a lot of space
14:40 phanimahesh Push to a git remote.
14:40 daniel_rose joined #git
14:40 phanimahesh And disable gc on server.
14:40 phanimahesh You'll have as good a backup as you can get.
14:41 CalimeroTeknik push from what?
14:41 CalimeroTeknik what about the gitignored files?
14:41 rantic_work joined #git
14:42 CalimeroTeknik and it's not one but dozens of repos that are in there
14:42 johnny56_ joined #git
14:43 CalimeroTeknik I know it's a mindset problem, but from your answer I can only derive one, even worse monstrosity: since git won't add all files, I can follow tarballs of that all these files with git
14:43 CalimeroTeknik which is highly inconvenient
14:43 kbs joined #git
14:44 phanimahesh Oh. so you have multiple git repos, and you want to have backups to be able to restore to the exact state, including ignored files.
14:44 CalimeroTeknik yeah, I could do tarballs every day, and even deal with the trouble of making incremental tarballs
14:44 CalimeroTeknik but git sounded like a cool revision history tool, except it refuses to follow everything
14:44 ribasushi joined #git
14:45 phanimahesh Is some scripting an option?
14:45 phanimahesh The contents of .git are not supposed to be version controlled. Trying to do that will end in a monstrosity worse than tarballs.
14:45 CalimeroTeknik a tailor-made solution? that's more likely to leave some stuff out than a bulky solution
14:45 GodGinrai CalimeroTeknik: dozens of what kind of repos? git repos?
14:45 CalimeroTeknik yes
14:46 GodGinrai Why not just push all of those repos to another server
14:46 phanimahesh Two .git directories can be equivalent even when they are wildly different from the view of file system.
14:46 CalimeroTeknik what about ignored files? (data files)
14:46 phanimahesh Have you seen !config
14:46 gitinfo [!configfiles] It is recommended to store local configuration data in a file which is not tracked by git, but certain deployment scenarios(such as Heroku) may require otherwise. See https://gist.github.com/1423106 for some ideas
14:46 ToxicFrog CalimeroTeknik: if you want a backup/sync tool that uses git as the backend/transport format, there are a bunch; bup and sparkleshare come to mind
14:46 GodGinrai CalimeroTeknik: if they are ignored, then they don't belong in the repo in the first place, so you shouldn't need to back them up.
14:47 CalimeroTeknik they don't belong in the repository for sure, but when I backup the whole thing I want them in
14:47 GodGinrai CalimeroTeknik: why, what are they?  Build Artifacts?
14:47 CalimeroTeknik config and dictionary data
14:48 CalimeroTeknik the code in the multiple repos is the application code for accessing the dictionary, or using that, etc
14:48 stantonk joined #git
14:48 bremner the thing about backups is you don't want them to be a special hack only you understamd
14:48 CalimeroTeknik absolutely
14:48 CalimeroTeknik thanks for the pointers, I'll see if bup/sparkleshare can do that for me
14:48 bremner so just use a "normal" backup software, there are many choices
14:48 GodGinrai I think you should just backup with rsync or something
14:48 lindenle joined #git
14:49 GodGinrai rather than trying to shoehorn git
14:49 CalimeroTeknik `rsync log` doesn't exist and makes no sense, right?
14:49 CalimeroTeknik git log does
14:49 GodGinrai No, your comment just now made no sense
14:49 GodGinrai if you want a log of backups, you can easily make that
14:50 GodGinrai git is a source control tool, not a backup tool
14:51 ToxicFrog GodGinrai: rsync is a sychronization tool but doesn't really make a good backup tool, although you can hack together one on top of it
14:51 ToxicFrog CalimeroTeknik: FWIW I've been using bup in production for a few years now, storing 2-300GB of backed up data, without issues
14:51 GodGinrai ToxicFrog: it makes for a better one than git does.
14:51 bremner sure. But there are still _many_ backup tools.
14:51 ShekharReddy joined #git
14:51 ToxicFrog GodGinrai: oh, certainly
14:51 baraba joined #git
14:51 bremner duplicity, obnam, bup, ...
14:51 maaarghk rsync [...] > /some/git/repo
14:52 maaarghk if you must
14:52 maaarghk but i agree with everyone else just use a proper backup package
14:54 leeN joined #git
14:56 clemf joined #git
14:56 CalimeroTeknik ToxicFrog, do you get functionality equivalent to git diff / git log with bup?
14:57 CalimeroTeknik being able to see line by line diffs between snapshots
14:57 grawity wonder if you could 'subtree-merge' all the repos you're trying to back up
14:58 ToxicFrog CalimeroTeknik: "git log", yes. "git diff" only if your backups are small enough for git to compute the diff without crashing.
14:58 ToxicFrog In my case they aren't.
14:59 ToxicFrog (you can access bup repos with normal "git" tools, but a lot of git commands don't work well when the repo is hundreds of GB in size)
14:59 CalimeroTeknik nice! I'm not backing up more than a few hundred MBs anyway
15:00 tvon joined #git
15:02 ferr joined #git
15:02 drewo joined #git
15:03 GodGinrai grawity: he's trying to backup non-tracked files as well
15:04 catag87 joined #git
15:05 SwiftMatt joined #git
15:05 rgrinberg joined #git
15:06 diegoviola joined #git
15:08 ferr joined #git
15:09 pcre_ joined #git
15:11 maaarghk left #git
15:12 GodGinrai joined #git
15:12 GodGinrai joined #git
15:13 GT4066 joined #git
15:14 chachasmooth joined #git
15:15 Masterphi joined #git
15:16 pcre_ joined #git
15:19 King_DuckZ in my script I want to get the full absolute remote address to a git submodule, in a way that git clone understands - so if current branch is tracking origin/master and origin is github:KingDuckZ/blah.git, and the submodule's path is ../foo.git, I want to get github:KingDuckZ/foo.git
15:19 King_DuckZ I'm not sure how to handle that in my ruby script, so I'm wondering if git can already provide that info
15:20 zivester joined #git
15:22 DieguezZ joined #git
15:23 MZAWeb joined #git
15:24 romerocesar joined #git
15:24 gkatsev joined #git
15:25 ash_workz joined #git
15:25 stuh84 joined #git
15:26 Asiajey joined #git
15:27 tristanp joined #git
15:28 hahuang61 joined #git
15:30 Rodya_ joined #git
15:31 lugzer joined #git
15:32 ajf- joined #git
15:34 romerocesar joined #git
15:36 hmdkny joined #git
15:36 cdown joined #git
15:36 lindenle joined #git
15:36 shinnya joined #git
15:38 chrisshattuck joined #git
15:38 TheodoreBe joined #git
15:40 InfoTest joined #git
15:41 Masterphi joined #git
15:41 stantonk joined #git
15:45 thiago joined #git
15:46 EvilPenguin joined #git
15:46 kpease joined #git
15:49 dreiss joined #git
15:49 Passe_a_Manteiga left #git
15:51 cdown joined #git
15:54 crameth joined #git
15:55 cxybeb joined #git
15:55 dmj` joined #git
15:55 D[4]ni joined #git
15:56 hashpuppy joined #git
15:57 UncleCJ joined #git
15:57 drewo joined #git
15:58 ArchNoob joined #git
15:58 elosz joined #git
15:58 spriz joined #git
16:00 SteffanW joined #git
16:00 stantonk joined #git
16:02 acetakwas joined #git
16:04 sangy joined #git
16:06 acetakwas joined #git
16:06 circ-user-foXmx joined #git
16:08 PabloR joined #git
16:10 jimi_ joined #git
16:11 daniel_rose joined #git
16:12 arooni_ joined #git
16:13 zeroed joined #git
16:13 madewokherd joined #git
16:17 xlegoman joined #git
16:17 arooni_ left #git
16:18 Ranhir joined #git
16:18 ImJune joined #git
16:19 arooni joined #git
16:21 NeXTSUN joined #git
16:22 stantonk joined #git
16:24 romerocesar joined #git
16:24 emPi joined #git
16:27 PioneerAxon joined #git
16:29 tristanp joined #git
16:29 hahuang61 joined #git
16:29 dmto joined #git
16:30 PioneerAxon joined #git
16:32 encod3 joined #git
16:32 lugzer joined #git
16:32 ascarter joined #git
16:32 stantonk joined #git
16:33 preyalone joined #git
16:34 Matrixiumn joined #git
16:35 DieguezZ joined #git
16:36 catag87 joined #git
16:37 IWishIKnew joined #git
16:38 OnBrokenWings joined #git
16:39 jwarren_ joined #git
16:39 texinwien joined #git
16:40 mrkat- joined #git
16:41 jwarren_ Howdy folks. I'm migrating a heafty repo from GitHub to Bitbucket. Of course migrating the code is easy, but I'm also trying to migrate closed and open pull requests, which seems to be more difficult. I found https://gist.github.com/piscisaureus/3342247 which describes how to pull PRs to your local machine, but is it possible to push PRs to a remote? IE the new Bitbucket repo? I'm guessing that the answer is no, and I probably need t
16:41 jwarren_ Thoughts?
16:42 PabloR I have git 2.6.2 on my OSX 10.9.5 Somebody helped me and I don’t recall installation method. Should I update thru homebrew or other method is preferable? Should I delete Git and start from scratch?
16:42 dangerousdave joined #git
16:42 jwarren_ I prefer to update through Homebrew, but I know others don't care for that method.
16:43 drewo joined #git
16:43 PabloR If homebrew wasn't the original method, does it matter if I use it now? Should I delete git altogether and start from scratch?
16:46 cdg joined #git
16:47 mdw joined #git
16:48 TheVoid- joined #git
16:49 jwarren_ PabloR If homebrew wasn't the original installation, it'll just install a new Git and add it to the system path, I believe. That's the way I prefer to do it, I don't like writing over system default stuff generally.
16:49 tango_ joined #git
16:49 PabloR jwarren_  ok thanks
16:50 pioiytr joined #git
16:50 jwarren_ I don't know Homebrew super well, but there's a way you can look up what's been installed with it. Try brew --help or something.
16:50 PabloR great. thanks
16:51 TomyLobo joined #git
16:51 jwarren_ Looks like the command is "brew list" @PabloR
16:52 PabloR Ah, git came up :p
16:52 jimi_ joined #git
16:52 jwarren_ PabloR: Then I think you can do something like "brew upgrade git".
16:53 PabloR jwarren_  Yes. I got to this page https://github.com/Homebrew/brew/blob/master/share/doc/homebrew/FAQ.md
16:55 thiago joined #git
16:55 romerocesar joined #git
16:56 jwarren_ PabloR: Awesome, good luck with the upgrade!
16:56 PabloR it's upgrading. Thanks
16:57 cyan__ joined #git
16:58 nwkj86 joined #git
17:01 Masterphi joined #git
17:02 aguimaraes1986 joined #git
17:03 freimatz joined #git
17:03 stantonk joined #git
17:04 cdown_ joined #git
17:04 jwarren_ left #git
17:06 blackwind_123 joined #git
17:08 King_DuckZ left #git
17:08 usm4n joined #git
17:08 tristanp joined #git
17:09 OnBrokenWings joined #git
17:09 stuh84 joined #git
17:10 mariuscc joined #git
17:12 Rodya_ joined #git
17:12 kbs joined #git
17:12 stantonk joined #git
17:14 tristanp_ joined #git
17:16 dunpeal joined #git
17:17 irco joined #git
17:18 dunpeal How do I control the "Committer Name" field of a commit from the command line?  I can control the Author with --author, is there anything similar for the committer name?
17:18 nnyk_ joined #git
17:19 Doginal joined #git
17:20 pioiytr joined #git
17:21 dimi1947 joined #git
17:23 EvilPenguin joined #git
17:23 pwnz0r joined #git
17:23 NeXTSUN joined #git
17:23 sfjohansson joined #git
17:24 mathewdgardner joined #git
17:25 sfjohansson Hey guys, I would like to get rid of alot of old history in my git repo, I’m using sourcetreee as my client
17:25 mehola joined #git
17:25 ojacobson Read as: "I would like to author a completely new, shorter history"
17:25 romerocesar joined #git
17:25 ojacobson because that's the only way to do that, in Git.
17:25 usm4n joined #git
17:27 sfjohansson sounds like that would be it :)
17:27 OnBrokenWings joined #git
17:28 durham joined #git
17:28 manizzle joined #git
17:28 manizzle hey guys. I do git remote add $BLAHBLAH
17:28 manizzle but I only want to add it for fetchs, not push
17:29 manizzle I dont want to accidentally push to that remote ever
17:29 manizzle http://stackoverflow.com/questions/7556155/git-set-up-a-fetch-only-remote
17:29 manizzle nvm found
17:30 hahuang61 joined #git
17:31 GodGinrai manizzle: you can use quotes to make that url something more legible
17:31 GodGinrai so that you get an error message that makes sense
17:31 GodGinrai as seen here: http://stackoverflow.com/questions/10270027/can-i-mark-a-git-remote-as-read-only
17:31 Silenced_v2 joined #git
17:32 PabloR jwarren_ or anybody. I had an issue with updating git thru homebrew. Can anybody help?
17:32 PabloR upgrade: Error: The `brew link` step did not complete successfully. The formula built, but is not symlinked into /usr/local. Could not symlink bin/git”.
17:32 manizzle GodGinrai, thanks!
17:32 TheodoreBe joined #git
17:32 bremner joined #git
17:32 bremner joined #git
17:33 PabloR Seems that I had git installed with 2 different methods because when I first run “git version” gave me back 2.6.2. However brew list gave me 2.4.3 and upgraded to 2.9.1. There was an error during the brew upgrade: Error: The `brew link` step did not complete successfully. The formula built, but is not symlinked into /usr/local. Could not symlink bin/git”.
17:33 circ-user-foXmx joined #git
17:33 lugzer joined #git
17:33 troulouliou_div2 joined #git
17:33 Silenced_v2 joined #git
17:34 Rodya_ joined #git
17:34 hahuang65 joined #git
17:35 PabloR I have a bit more info. I really appreciate the help
17:35 kbs dunpeal: dunno about args, but GIT_COMMITTER_NAME='a' GIT_COMMITTER_EMAIL='b' git commit -m 'foo' should work ?
17:36 rgrinberg joined #git
17:36 mwhooker joined #git
17:36 usm4n joined #git
17:36 blackwind_123 joined #git
17:38 FunkyAss joined #git
17:40 TheodoreBe joined #git
17:40 kulelu88 joined #git
17:41 PioneerAxon joined #git
17:43 tristanp joined #git
17:43 sfjohansson ojacobson what would be the workflow for writeing a new shorter history?
17:44 bronson joined #git
17:45 anth0ny joined #git
17:46 snowkidind joined #git
17:46 dunpeal kbs: thanks.
17:46 snowkidind I have an xcode project which contains two targets: a osx shell and a static library
17:46 Repox joined #git
17:46 snowkidind what is the recommended repo structure for something like that
17:47 snowkidind (i'm developing the library from within the workspace)
17:48 dunpeal Github question: I'm trying to create a simple PR, but when I review it, I see changes in addition to the ones I made, specifically: CHANGELOG.md is modified. How do I control/prevent these additional changes?
17:48 ojacobson sfjohansson: I'd start with man git-filter-branch personally, but there's no easy way to do it
17:48 gitinfo sfjohansson: the git-filter-branch manpage is available at http://jk.gs/git-filter-branch.html
17:48 ojacobson dunpeal: that change must have been introduced by some commit in `git log origin/target-branch...your-branch`
17:48 ojacobson (three dots)
17:48 ojacobson Rewrite your branch to not include that change (and probably put it on another branch)
17:50 PabloR Any help with my git installation issue?
17:50 * kbs wonders if git commit-tree is the crude answer for sfjohansson :-)
17:53 sfjohansson hmm… so I’ve been using git quite basically… what is the idea cocnept with these rewriting of history…how should I look at it?
17:54 sfjohansson so now i prettymuch have jsut one ver very long branch
17:55 sfjohansson do I create a new bracnh and sort of copy over “grouped commits” and moving back and forth deleting etc… =O
17:55 preaction PabloR: that's a homebrew issue, not a git issue. look for help from the homebrew folks
17:56 PabloR preaction ok thanks
17:56 kbs sfjohansson: the underlying data model is very simple - a commit is literally a data structure with a pointer to a tree, and zero or more commit objects (its parents.) You can invent a new commit object (for instance) that points to your latest tree and no parent, and this will look like a branch with zero history, but with all the files of interest.
17:57 Sasazuka joined #git
17:57 kbs there aren't really any semantics about a commit other than this - all git does is provide (endlelly confusing?) ways to manipulate them ;-)
17:58 usm4n joined #git
17:58 sfjohansson kbs it certianly seem to work on confusing me…..=O
17:58 dunpeal ojacobson: It doesn't look like it. I just forked origin, made a simple change to master, and now trying to merge back to origin/master.
17:58 dunpeal It looks like some sort of automated change...
17:59 kbs sfjohansson: haha. I've found much zen in just doing "git cat-file -p HEAD" and looking at what it means
17:59 dunpeal autogenerated by the pull request?  Is that something Github supports?
17:59 mathewdgardner joined #git
18:00 LionsMane joined #git
18:00 johnmilton joined #git
18:00 Silenced_v2 joined #git
18:00 Silenced_v2 joined #git
18:01 ojacobson No.
18:02 deBase joined #git
18:02 stantonk joined #git
18:02 sfjohansson kbs hmm..at least I odnt feel closer on how to solve this this…. =O
18:03 azbarcea joined #git
18:04 davisonio joined #git
18:05 dunpeal ojacobson: d'oh. I amended a remote commit by mistake :/
18:06 kaldoran joined #git
18:06 romerocesar joined #git
18:08 SwiftMatt joined #git
18:08 realz joined #git
18:11 mikecmpbll joined #git
18:12 daniel_rose joined #git
18:13 daniel_rose joined #git
18:14 kbs sfjohansson: dunno if there's a cookbook tutorial that does exactly what you want. But should you feel that (essentially) you want to invent a commit history that picks selected states of the code in some suitable way, with an invented commit messages - a crude hint might be http://paste.debian.net/hidden/77c2ac19/ but it's probably needs understanding what git does at the low level
18:14 kbs (that said, I personally think the only way git makes sense to me is at the low level :)
18:14 kbs s/git makes sense to me/I make sense of git/
18:15 cuebix-wk joined #git
18:17 johnmilton joined #git
18:18 Doginal joined #git
18:20 sfjohansson kbs I guess thats always the thing with using a GUI for anything =O now it feelsl ike my old solution with just deleting the repo and initializing a new one was pretty straightfroward but I can’t relly do it this time  as I already have a few commits…that I need to keep =(
18:22 quetz joined #git
18:22 kbs *nod* fwiw, I sorta slowly came realize that dealing with git is actually considerably easier without the guis, and spending an hour or two figuring out the low-level bits (which is actually fairly simple.) Then, the guis and commands made much more sense to me, and I could make git work for me (than vice-versa :)
18:23 jost__ joined #git
18:23 SwiftMatt joined #git
18:23 tristanp_ joined #git
18:24 sfjohansson kbs sounds like the better approach…and is more cross platofrm :)
18:24 rwb joined #git
18:25 _ikke_ kbs: could not agree more
18:25 stantonk joined #git
18:26 mathewdgardner joined #git
18:26 ArchNoob joined #git
18:27 dmj` joined #git
18:28 Rodya_ joined #git
18:28 jason237 joined #git
18:28 Timvde joined #git
18:30 tristanp joined #git
18:30 Silenced_v2 joined #git
18:30 Silenced_v2 joined #git
18:31 xall joined #git
18:31 mariuscc joined #git
18:32 TheSpiderPig joined #git
18:32 hck89 joined #git
18:32 Rodya_ joined #git
18:32 drewo joined #git
18:33 yohnnyjoe joined #git
18:33 pcre_ joined #git
18:34 ZenWalker joined #git
18:34 ZenWalker hello, how to search a commit ignoring the caps?
18:35 lugzer joined #git
18:35 i7c git log -i --grep='pattern'
18:36 PabloR left #git
18:36 d4rklit3 joined #git
18:36 mikecmpbll joined #git
18:36 TheSpiderPig grep takex a regex
18:36 TheSpiderPig afaik
18:36 i7c yup
18:37 i7c But you'll have a hard time doing --grep='[tT]his [iI]s [wW]hat [iI]m [lL]ooking [fF]or' :)
18:37 i7c Dunno if it understands some case insensitive directive like some regex do.
18:37 pioiytr_ joined #git
18:38 TheSpiderPig cant you use a /i
18:38 TheSpiderPig it *is* regex
18:38 f3r70rr35f joined #git
18:38 TheSpiderPig oh fuck, yeah use -i
18:38 TheSpiderPig or --ignore-case
18:39 TheSpiderPig wait that's for gitr grep
18:39 i7c So --grep='/something/i' does not work if that’s what you meant.
18:41 guampa joined #git
18:41 Literphor joined #git
18:43 Masterphi joined #git
18:43 sargas joined #git
18:43 stonerfish joined #git
18:44 TheSpiderPig left #git
18:44 mathewdgardner joined #git
18:44 mutex joined #git
18:46 Rodya_ joined #git
18:50 SwiftMatt joined #git
18:51 Silenced_v2 joined #git
18:51 sfjohansson kbs Im ight have done it…at least it looks like it…but maybe everything is there but hidden =O
18:51 Silenced_v2 joined #git
18:52 ekinmur joined #git
18:52 l4v2 joined #git
18:53 chachasmooth joined #git
18:53 sfjohansson ojacobson I did try a graft and git filter-branch ..it looks fine in my git client…would this have actually thrown away history too or is it just “represented” in cleaner way
18:54 ojacobson Without seeing what you actually did, I have no idea.
18:54 Sasazuka__ joined #git
18:54 ojacobson (That wasn't an invitation; I'm busy right now, but if you paste a !transcript I'm sure someone can add commentary. :)
18:54 gitinfo Please paste (using https://gist.github.com/ or https://vomitb.in/ ) a transcript ( https://gist.github.com/2415442 ) of your terminal session so we can see exactly what you see
18:55 tvon joined #git
18:55 bronson joined #git
18:55 jungsubk joined #git
18:57 mischat joined #git
18:58 pioiytr__ joined #git
18:58 p4trix joined #git
18:58 mwhooker joined #git
19:00 NeXTSUN joined #git
19:00 ZenWalker -i works, thanks
19:00 ZenWalker left #git
19:01 sfjohansson ojacobson Cheers…here is the link http://hastebin.com/letewitifo.vbs :)
19:01 drewo joined #git
19:01 catag87 joined #git
19:01 circ-user-8eXFL joined #git
19:02 catag87 joined #git
19:03 EvilPenguin joined #git
19:03 tristanp joined #git
19:05 spudowiar joined #git
19:05 blackwind_123 joined #git
19:08 Waskeend joined #git
19:09 Repox joined #git
19:11 timmu joined #git
19:12 govg joined #git
19:15 paul424 joined #git
19:15 dunpeal joined #git
19:17 vF3hNGxc47h8 joined #git
19:18 UncleCJ joined #git
19:19 ynanm joined #git
19:20 nnyk_ joined #git
19:20 NeverDie joined #git
19:20 astrofog joined #git
19:23 Rodya_ joined #git
19:27 chachasmooth joined #git
19:27 troulouliou_div2 joined #git
19:28 muelli joined #git
19:28 editshare-ericf joined #git
19:29 BlaXpirit I want to see whether I can push to a remote without needing -f.  is there something like --dry-run and will it do this?
19:29 troulouliou_div2 joined #git
19:31 weox joined #git
19:31 BlaXpirit nevermind, I don't really need to check, I can just do it
19:31 Darcidride joined #git
19:32 hahuang61 joined #git
19:32 invisbl joined #git
19:33 timmu joined #git
19:34 stantonk joined #git
19:34 jungsubk joined #git
19:36 lugzer joined #git
19:36 jimi_ joined #git
19:36 Rodya_ joined #git
19:37 satifant joined #git
19:37 l4v2 joined #git
19:38 slayerjain joined #git
19:38 stantonk joined #git
19:38 tristanp_ joined #git
19:39 hunter2 joined #git
19:40 mwhooker joined #git
19:41 hunter2 hello all. i have much to learn. can someone tell me how you identify a file by its revision number?
19:41 ojacobson files don't have revision numbers in git
19:42 ojacobson there's a hash for each distinct blob (file contents), but the same blob can appear in multiple paths or in multiple revisions, so asking "what's the path to this blob" is a bit complex
19:43 elastix joined #git
19:43 Lynxium joined #git
19:44 Bennux joined #git
19:44 hunter2 supposing i commit a.txt and it's hash is 4f9542d.. how do i show that revision?
19:45 preaction git show 4f9542d
19:45 hunter2 tried, a@4f... a@{4f} ... i'm missing something
19:45 mikecmpbll joined #git
19:45 hunter2 i just want a.txt though
19:45 djensen joined #git
19:45 i7c git show 4f9542d:a.txt
19:45 preaction git show 4f9542d -- a.txt
19:45 hunter2 oh ffs, it goes after
19:46 spudowiar bloob
19:46 i7c preaction, that shows the diff.
19:46 spudowiar blobby bob
19:46 rahtgaz joined #git
19:46 spudowiar i7c: well, that's what that blob is :)
19:47 EvilPenguin joined #git
19:47 i7c Not what hunter2 asked, if I understood him right tho.
19:47 shuaib joined #git
19:47 GodGinrai joined #git
19:47 spudowiar well, yeah
19:47 spudowiar but :)
19:47 spudowiar I'm just here to trollll
19:48 spudowiar don't mind me
19:48 i7c spudowiar, besides that’s wrong.
19:48 hunter2 another question, suppose i'm going round adding a copyright boiler plate, i commit that to a.txt, how do i take that commit's diff and apply that to b
19:48 i7c The blob will contain the entire file not just the diff.
19:48 djensen left #git
19:48 spudowiar i7c: but that'd be inefficient o_O
19:48 i7c Not really.
19:49 spudowiar well, if you add 1 line to a 100MB file, it'd be two 100MB blobs?
19:49 i7c At first, yes. At some point they'll be delta compressed and whatnot.
19:50 ojacobson spudowiar: git's logical model is purely snapshot based, so, initially, yes - there will.
19:50 spudowiar :O
19:50 spudowiar TIL
19:50 ojacobson however, git optimizes that by delta-encoding similar objects when packing loose objects
19:50 spudowiar thanks
19:50 ojacobson The encoded deltas have nothing to do with commit diffs, exactly - it's more like "find me the most similar, and diff from that"
19:51 i7c And it's not really "inefficient" that way. If it all just were diffs you’d have to apply diffs for hours if the history is long enough.
19:52 sargas joined #git
19:53 TheodoreBe joined #git
19:54 GodGinrai joined #git
19:54 mwhooker joined #git
19:54 pcre_ joined #git
19:56 hunter2 so, i was thinking something like this: git merge-file b.txt 4f1af:a.txt 8a8a8a:a.txt
19:56 lugzer joined #git
19:58 usm4n joined #git
19:59 hunter2 ... but that doesn't work. i've much to learn so any pointers would be appreciated!
20:01 Silenced joined #git
20:01 Masterphi joined #git
20:02 synthroid joined #git
20:02 thiago joined #git
20:04 Asiajey joined #git
20:04 Asiajey joined #git
20:04 rpd hunter2: what error does that give?
20:06 rpd I guess merge-file doesn't take a commit-ish argument and doesn't resolve the file
20:06 Cooler1 joined #git
20:06 hunter2 Could not stat 4f954...:a
20:07 hunter2 which is fair enough, its not on disk as that name .. but was hoping git would work it out as a point in the repos
20:07 rpd yeah, well... you could do something like git cat-file ... > a_file etc
20:07 rpd and then use merge file to merge the results of those cat-file operations
20:08 alexruf_ joined #git
20:08 tristanp joined #git
20:08 pcre_ joined #git
20:10 rpd yeah, the user-manual gives that as an example for using merge-file to merge files from different trees
20:10 ArchNoob joined #git
20:10 alexruf joined #git
20:11 elastix joined #git
20:12 chachasmooth joined #git
20:12 yqt joined #git
20:12 fus joined #git
20:12 usm4n joined #git
20:14 tristanp joined #git
20:14 pcre joined #git
20:14 DeJoker joined #git
20:14 DeJoker Test ?
20:15 DeJoker Any knowledgable folks here ?
20:15 selckin just jokers
20:15 DeJoker Great :)
20:16 preaction also space cowboys. but no gangsters of love
20:16 GodGinrai and some battle programmers
20:17 romerocesar joined #git
20:17 lugzer joined #git
20:17 mariuscc2 joined #git
20:18 preaction what about Maurice?
20:18 DeJoker Okay I have a Source.pgm in git — that another team works on — and I have some changes that I make to that Source.pgm to turn it into Ultimate.pgm which changes a few lines in a few files, deletes a few files, replaces a few files, and add some new files — can I have a repository of just those alterations that I can then Merge into the Source.pgm to make my Ultimate.pgm ??  Or must I regrettably have all the content in my development repository ???
20:19 MaximusColourum joined #git
20:20 fedya___ joined #git
20:21 fus joined #git
20:21 fedya___ Hi, I made a huge mistake just now. I meant to commit to a new branch, but accidentally committed to master and pushed to remote. I reverted the commit with "git reset HEAD^ --hard" and "git push origin -f" –– now all my work is gone... I wasn't thinking... Is there anyway to restore the commit?
20:22 Waskeend joined #git
20:23 preaction fedya___: check the git reflog
20:23 preaction or if you remember the sha of the commit, it's likely still there
20:23 Cease joined #git
20:28 dmj`` joined #git
20:28 wagle joined #git
20:28 DeJoker Okay I have a Source.pgm in git — that another team works on — and I have some changes that I make to that Source.pgm to turn it into Ultimate.pgm which changes a few lines in a few files, deletes a few files, replaces a few files, and add some new files — can I have a repository of just those alterations that I can then Merge into the Source.pgm to make my Ultimate.pgm ??  Or must I regrettably have all the content in my development repository ???   Not
20:28 DeJoker looked into SubModules but this did not seem to be the correct tool but then again maybe I did not look deep enough as of yet
20:29 blackwind_123 joined #git
20:31 DeJoker Hmmm okay just Jokers, Smokers and Tokers I guess :)  as I was told
20:31 usm4n joined #git
20:32 danimal2 joined #git
20:32 muelli joined #git
20:33 texinwien joined #git
20:33 fedya___ preaction: I looked at .git directory and it looks like it's still there:
20:33 fedya___ e4b5738 HEAD@{0}: reset: moving to HEAD^
20:33 fedya___ 6ca57e2 HEAD@{1}: commit: [PLAT-2832] Daily cronjob at 10am to send final changes notice for orders shipping in 45 days
20:33 fedya___ how do I restore it? and put it in a new branch? I currently have master checked out
20:33 mand0 joined #git
20:34 preaction !reflog
20:34 gitinfo The git reflog (`git log -g`) records the SHAs of your HEADs for 2+ weeks. `git checkout -b myrestore OLDSHA` and `git reset --hard OLDSHA` will relink to that state via a new and current branch respectively, see http://sethrobertson.github.com/GitFixUm/ for full details.  WARNING: reset --hard will trash any uncommitted changes!  Visualize with: gitk --all --date-order `git log -g --pretty=%H`
20:36 mathewdgardner joined #git
20:36 fedya___ sorry, I'm a bit confused, worried about doing another hard reset. I'm following the part with checking out into new branch using the SHA I grabbed from reflog, what will the git reset --hard OLDSHA do?
20:37 stantonk_ joined #git
20:38 neurodrone_ joined #git
20:38 lugzer joined #git
20:38 stantonk_ joined #git
20:40 preaction fedya___: those are two different things. one makes a new branch, the other changes the current branch
20:40 preaction see the link
20:43 fedya___ there's only one file with M flag there, there should be 4...
20:43 fedya___ could I have lost that commit completely since I did a reset --hard?
20:45 bronson joined #git
20:45 preaction no
20:46 dunpeal joined #git
20:47 TheodoreBe joined #git
20:47 DeJoker Okay I have a Source.pgm in git — that another team works on — and I have some changes that I make to that Source.pgm to turn it into Ultimate.pgm which changes a few lines in a few files, deletes a few files, replaces a few files, and add some new files — can I have a repository of just those alterations that I can then Merge into the Source.pgm to make my Ultimate.pgm ??  Or must I regrettably have all the content in my development repository ???  Not
20:47 DeJoker looked into SubModules but this did not seem to be the correct tool but then again maybe I did not look deep enough as of yet  — So nobody can help me with this ??
20:48 TheodoreBe joined #git
20:49 nnyk joined #git
20:49 mwhooker joined #git
20:51 Kulrak joined #git
20:51 Kulrak joined #git
20:51 DeJoker left #git
20:54 DeJoker joined #git
20:55 Sadeukko joined #git
20:55 DeJoker left #git
20:55 DeJoker joined #git
20:56 DeJoker left #git
20:57 drewo joined #git
20:57 white_gecko joined #git
20:58 drewo joined #git
20:58 Doginal joined #git
20:59 pioiytr_ joined #git
21:01 Rodya_ joined #git
21:01 FuzzySockets joined #git
21:02 tristanp joined #git
21:03 Silenced joined #git
21:04 nilg joined #git
21:04 sapiens_vector joined #git
21:05 daniel_rose joined #git
21:05 Silenced joined #git
21:07 jagob joined #git
21:08 Orion3k joined #git
21:08 Sucks joined #git
21:10 wagle joined #git
21:11 NeonSynth_ joined #git
21:17 xall joined #git
21:18 tgunr joined #git
21:19 ajf- joined #git
21:21 bunniez joined #git
21:22 bunniez hello friends, i need help unsetting an alias. somehow the alias in question("ci") got defined twice, shows up twice in the list "git config --get-regexp alias". when i run "git config --global --unset alias.ci", i get the message "warning: alias.ci has multiple values" and it does not do it. how do i force it to remove this alias?
21:25 rpd bunniez: you can't manually remove it?
21:25 diogenese joined #git
21:26 mwhooker joined #git
21:27 bunniez rpd: i thought that "git config --global --unset alias.ci" was the way to do that. is there another way?
21:28 rpd I mean, if it's giving you an error because it's defined twice, just edit the file and remove both references to make that command work again
21:28 bunniez rpd: thanks, sounds like that would work. where do i locate the file containing global aliases?
21:28 rpd ~/.gitconfig probably
21:29 bunniez rpd: thank you. i'm on a windows machine, do you figure ~ would probably be the C:/ drive?
21:29 mathewdgardner joined #git
21:31 rpd For windows I'm not sure. You can go look at what HOME is defined as in C:\Program Files\Git. It's probably something like C:\Users\Something\.gitconfig
21:31 ahmed_elgabri joined #git
21:31 UncleCJ joined #git
21:31 ShekharReddy joined #git
21:31 synthroid joined #git
21:33 bunniez rpd: you're right, it was in Users, thanks!!
21:34 hahuang61 joined #git
21:34 Dreamer3 joined #git
21:34 bruce_lee joined #git
21:34 bruce_lee joined #git
21:35 bunniez left #git
21:36 stantonk joined #git
21:36 tristanp joined #git
21:37 spudowiar joined #git
21:37 mwhooker joined #git
21:38 SimonNa joined #git
21:41 EvilPenguin joined #git
21:42 aidalgol joined #git
21:43 usm4n_ joined #git
21:43 ahmed_elgabri joined #git
21:45 white_gecko joined #git
21:46 Brain joined #git
21:46 azbarcea joined #git
21:46 jason237 joined #git
21:48 Muzer joined #git
21:52 drewo joined #git
21:53 stantonk joined #git
21:53 f3r70rr35f joined #git
21:54 DieguezZ joined #git
21:54 Gsham joined #git
21:54 HoierM joined #git
21:55 pneuma joined #git
21:58 romerocesar joined #git
22:02 raijin joined #git
22:04 Literphor joined #git
22:05 joki joined #git
22:05 Masterphi joined #git
22:09 stonerfish joined #git
22:10 jimi_ joined #git
22:11 tristanp joined #git
22:11 Atrumx joined #git
22:13 nettoweb joined #git
22:15 Sucks joined #git
22:18 romerocesar joined #git
22:20 a_n_d_y joined #git
22:20 Guest55 joined #git
22:21 a_n_d_y i have a usage question. it feels like i'm being stupid
22:21 a_n_d_y will anyone help?
22:21 devbug joined #git
22:21 ojacobson Only one way to find out
22:21 a_n_d_y :)
22:22 Rodya_ joined #git
22:22 a_n_d_y I'm working with a few guys on a server using git. a lot of the time, changes need to be made live on the server. is there any way to all work ont he server at the same time, using git branches, without conflicting or committing directly to master all the time?
22:25 manizzle a_n_d_y, make a branch called dev. Everyone works in their own branches, merging against dev. Then 1 person is in charge of taking dev and merging with master
22:26 ojacobson "a lot of the time, changes need to be made live on the server" first things first: _no matter what source control tool you use_, this is going to cause bottlenecks.
22:26 EvilPenguin joined #git
22:26 ojacobson Make it possible to run at least a cut-down version of your service locally, or in docker, or in a VM, or something and you will fix a lot of this
22:27 a_n_d_y thanks manizzle, makes sense. like git flow or something i guess
22:27 a_n_d_y ojacobson, yeah it feels like we're working wrong
22:27 ojacobson Second, more specifically, no - git assumes a work tree belongs to a single workflow at a time.
22:28 ojacobson You could _in principle_ have multiple repos, but share a work tree, but it's not worth doing
22:28 a_n_d_y i dunno if we can create cutdown servers locally. there's so many it would take a lot of effort. would individual directories on the server for each user work?
22:28 Sucks joined #git
22:28 ojacobson you won't get what you want and you will get a bunch of headaches
22:28 a_n_d_y andy-dev dir
22:28 tristanp_ joined #git
22:28 mwhooker joined #git
22:30 bronson joined #git
22:31 Brain joined #git
22:35 tristanp joined #git
22:35 madewokherd joined #git
22:36 white_gecko left #git
22:37 dreiss joined #git
22:40 p4trix joined #git
22:40 Jellyg00se joined #git
22:41 Masterphi joined #git
22:42 FuzzySockets joined #git
22:42 magicbit joined #git
22:42 cd-rum joined #git
22:45 cd-rum_ joined #git
22:47 bronson joined #git
22:48 Vinnie_win joined #git
22:48 romerocesar joined #git
22:51 stantonk joined #git
22:51 mischat joined #git
22:53 xall joined #git
22:53 apotry joined #git
22:57 Ahmed90 joined #git
22:57 Ahmed-90 joined #git
22:57 barteks2x joined #git
22:59 Masterphi joined #git
23:00 drewo joined #git
23:01 tristanp joined #git
23:01 Masterphi joined #git
23:01 cdown joined #git
23:03 Literphor joined #git
23:04 nwkj86 joined #git
23:08 Masterphi joined #git
23:09 FuzzySockets joined #git
23:11 HanClinto joined #git
23:12 stantonk joined #git
23:14 m0viefreak joined #git
23:19 akiva joined #git
23:19 romerocesar joined #git
23:22 Cassiopaya joined #git
23:22 mathewdgardner joined #git
23:23 mrBen2k2k2k joined #git
23:23 cdown joined #git
23:23 nettoweb joined #git
23:25 sfjohansson_ joined #git
23:25 IlNominePatre joined #git
23:25 thiago joined #git
23:25 tristanp joined #git
23:28 anuxivm1 joined #git
23:29 Gamecubic joined #git
23:30 anuxivm1 left #git
23:32 anuxivm joined #git
23:32 stantonk joined #git
23:33 EvilPenguin joined #git
23:35 stantonk joined #git
23:38 rgrinberg joined #git
23:39 mrBen2k2k2k joined #git
23:41 synthroid joined #git
23:42 tristanp joined #git
23:42 catag87 joined #git
23:43 ravi joined #git
23:47 rahtgaz joined #git
23:49 romerocesar joined #git
23:50 Rodya_ joined #git
23:56 FuzzySockets joined #git
23:59 HanClinto joined #git

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